Reported Road Casualties Great Britain: 2008 Annual Report

Transcription

1 Reported Road Casualties Great Britain: 28 Annual Report

2

3 DEPARTMENT FOR TRANSPORT SCOTTISH GOVERNMENT WELSH ASSEMBLY GOVERNMENT Reported Road Casualties Great Britain: 28 Annual Report Published September 29 London: The Stationery Office

4 Department for Transport Great Minster House, 76 Marsham Street London SWP 4DR Telephone Internet service # Crown copyright 29 Copyright in the typographical arrangement rests with the Crown. This publication may be reproduced free of charge in any format or medium for non-commercial research, private study or internal circulation within an organisation. This is subject to it being reproduced accurately and not used in a misleading context. The material must be acknowledged as Crown copyright and the title of the document/publication specified. This publication can also be accessed at the Department s web site. For any other use of this material please apply for a Click-Use Licence at or by writing to OPSI at The Licensing Division, Office of Public Sector Information, St Clements House, 2 6 Colegate, Norwich NR3 BQ, Fax: or to licensing@cabinet-office.x.gsi.gov.uk. ISBN Printed in Great Britain in September 29 on material containing at least 75% recycled fibre. A National Statistics publication produced by Transport Statistics: DfT National Statistics are produced to high professional standards set out in the Code of Practice for Official Statistics. They undergo regular quality assurance reviews to ensure that they meet customer needs. They are produced free from any political interference. Contact Points: For general enquiries call the National Statistics Customer Contact Centre at: Room 5, ONS, Government Buildings, Cardiff Road, Newport, Gwent NP 8XG, Tel: , fax or info@statistics.gov.uk You can also obtain National Statistics through the internet go to For information relating to Transport Statistics go to Prepared for publication by: Anil Bhagat Linden Francis Pat Kilbey Barbara Noble Matthew Tranter David Wilson Christopher Waite Yingbo Xu DfT is often prepared to sell unpublished data. Further information and queries concerning this publication should be directed to: RSS, 2/2, Great Minster House, 76 Marsham Street, London SWP 4DR, Tel: +44 () , Fax +44 () , roadacc.stats@dft.gsi.gov.uk Acknowledgement The Department extends its grateful thanks to police forces and their officers for their contribution towards reducing road casualties, including the collection of STATS 9 data upon which this publication is based, and without which this government and road safety organisations would be much less well informed.

5 Introduction Reported Road Casualties Great Britain 28: Annual Report, (RRCGB) provides detailed statistics about the circumstances of personal injury road accidents, including the types of vehicles involved and the consequent casualties. Most of the statistics in this report are based on accidents reported to the police. There are also seven articles containing further analysis on specific road safety topics. These statistics are used to inform public debate and support policy on road safety. In 28, deaths in road accidents were less than half the number reported in the first edition of this report in 95. However, on average seven people were killed a day in 28. The collection and publication of detailed road safety data helps to identify future priorities to further reduce casualties. The Department for Transport has been working with the Scottish Government and the Welsh Assembly Government on a new road safety strategy for the period from 2. This will establish the approach, targets and measures for reducing casualties in the next decade and beyond. Consultation on the new strategy was carried out from April to July 29 and it is expected that the final strategy will be published by the end of 29. Very few, if any, fatal accidents do not become known to the police. However, it has long been known that a considerable proportion of non-fatal injury accidents are not reported to the police (and there is no legal obligation to do so). We have changed the titles of our publications to more closely reflect this. The Department is continuing to undertake research on levels of reporting. Article 5 in this publication discusses and compares other sources of data with police data on road casualties and provides a very broad estimate of the total number of road casualties. National and local government and police forces work closely to achieve an agreed national standard for the system for collecting and processing statistics on road accidents involving personal injury. The statistics are subjected to review about every five years as part of the continuing drive to improve quality and meet user needs whilst minimising the burden on providers. The external consultation process for the current review has now ended. The aim is for any changes to the system to be agreed this year and implemented in January 2. This year road casualty statistics have also been assessed by the United Kingdom Statistics Authority and retained their designation as National Statistics, subject to a number of recommendations to be met by November. Designation as National Statistics broadly this means that the statistics are considered to meet identified user needs; are produced, managed and disseminated to high standards; and are well explained. Pat Kilbey Head of Road Safety Statistics, DfT The full RRCGB report (in PDF format) and tables (in EXCEL format) are available from: Further information can be obtained from: Mr Anil Bhagat, telephone: , roadacc.stats@dft.gsi.gov.uk

6 Contents Page Articles. General overview and progress towards casualty reduction targets 6 2. A valuation of road accidents and casualties in Great Britain in Drinking and driving Contributory factors to road accidents 4 5. Comparing police data on road accidents with other sources An illustrative analysis of linked police and hospital data 8 7. Road Safety Research: An Overview 85 Notes 9 Notes to individual main tables Definitions, symbols and conventions Tables Part I Trends: average and 2 to 28 (unless stated) General a Vehicle population, traffic and road length: b Road traffic by vehicle type and road class: and average 2 Population, vehicle population, index of vehicle mileage, reported accidents and casualties: by road user type and severity: Accidents 3 Reported accidents and accident rates: by road class and severity 7 4 Reported accidents: by road class, speed limit and severity 8 Casualties 5a Reported male casualties: by built-up and non built-up roads, road class and severity 9 5b Reported female casualties: by built-up and non built-up roads, road class and severity 5c All reported casualties: by built-up and non built-up roads, road class and severity 6a Reported male casualties: by road user type and severity 2 6b Reported female casualties: by road user type and severity 3 6c All reported casualties: by road user type and severity 4 7a Reported male casualties: killed or seriously injured: by road user type and age 5 7b Reported female casualties: killed or seriously injured: by road user type and age 6 7c All reported casualties: killed or seriously injured: by road user type and age 7 8 Reported casualties: by time of accident and severity: Reported casualty rates: by road user type and severity: Vehicles and drivers involved Vehicles involved in reported accidents and involvement rates: by vehicle type and severity of accident: Breath tests and breath test failures: by drivers and riders involved in accidents:

7 Part II Detailed tables 28 (unless stated) General 2 Reported accidents, vehicles and casualties: casualties by severity: by road class, built-up and non built-up roads 2 3 Reported accidents and casualties: by severity, road type and speed limit 22 Accidents 4 Reported accidents: by severity, number of casualties involved, built-up and non built-up roads and road class 23 5a Reported accidents: by daylight and darkness, road surface condition, built-up and non built-up roads and severity 24 5b Reported casualties: by daylight and darkness, road surface condition, built-up and non built-up roads and severity 24 6a Reported accidents: by daylight and darkness, weather condition, built-up and non built-up roads and severity 25 6b Reported casualties: by daylight and darkness, weather condition, built-up and non built-up roads and severity 25 7 Reported accidents: by daylight and darkness, road surface condition, built-up and non built-up roads, speed limit and street lighting 26 8 Reported accidents: by daylight and darkness, lighting conditions, special conditions and carriageway hazards 27 9 Reported accidents: by junction type, built-up and non built-up roads and severity 27 2 Reported single vehicle accidents: by object hit off carriageway: built-up and non built-up roads and severity 28 2 Reported accidents: by number of vehicles involved, built-up and non built-up roads, road class and severity Reported accidents involving pedestrians and one vehicle: by severity and vehicle type 3 23a Reported accidents, vehicle user and pedestrian casualties: by combination of vehicles involved in urban areas 3 23b Reported accidents, vehicle user and pedestrian casualties: by combination of vehicles involved in rural areas 32 23c Reported accidents, vehicle user and pedestrian casualties: by combination of vehicles involved in all areas 33 Casualties 24 Reported casualties: by built-up and non built-up roads and motorways, severity and road user type Casualties in reported accidents involving vehicles of different types: by built-up and non built-up roads, road class and severity Reported casualty and accident rates: by urban and rural roads, road class, road user type, severity and pedestrian involvement Number of reported casualties: by accident and casualty severity and road user type Reported casualties and casualty rates: by month, road user type and severity 38 29a Reported casualties: by day, road user type and hour of day 39 29b Reported casualties: killed or seriously injured: by day, road user type and hour of day 4 29c Reported casualties: all days: by severity, road user type and hour of day 4 3a Reported casualties: by age band, road user type and severity 42 3b Reported casualties: by age band, road user type and severity: average 43 3

8 Reported casualty rates: by age band, road user type and severity Reported pedestrian casualties: location by age band and by severity Reported pedestrian casualties: by location, age, road crossing type and severity Reported casualties: by age, road user type and severity Reported casualties in cars: by severity, age, seating position, built-up and non built-up roads Drivers and vehicles involved 36 Reported school pupil casualties on journeys to and from school: by road user type, severity, gender and age 37 Reported breath tests and breath test failures: all drivers and riders involved, by day of week and time of day 38a Drivers in reported accidents: by gender, number injured, road user type and age 38b Drivers in reported accidents: by gender, number injured, road user type and age: average 39 Reported breath tests and breath test failures: by road user type and age 4 Vehicles involved in reported accidents: by accident severity and vehicle type 4a Vehicles involved in reported accidents: by vehicle type, built-up and non built-up roads, road class and accident severity 4b Vehicles involved in reported accidents: by vehicle type, built-up and non built-up roads, road class and accident severity: average 42 Vehicles involvement rates for reported accidents: by vehicle type, urban and rural roads, road class, accident severity and traffic 43 Vehicles involved in reported accidents: by junction type, vehicle type, built-up and non built-up roads 44 Vehicles involved in reported accidents skidding or overturning, and towing: by road surface condition, special conditions at site and vehicle type 45 Vehicles involved in reported accidents: by vehicle type and manoeuvre Area comparisons 46a Reported casualties: by road user type, severity and local authority 62 46b Reported casualties: by road user type, severity and local authority: average Reported casualties: by Government Office Region, country and severity: average, Casualties: by built-up and non built-up roads, road class, Government Office Region and severity 7 United Kingdom 49 Reported casualties: by severity, road user type and country: United Kingdom 72 Mortality 5 Deaths: by age and gender, from all causes, all accidental deaths and all road deaths: International comparisons 5 International comparisons of road deaths: number and rates for different road users: by selected countries: 27 and 28 (provisional) 74 Intermodal comparisons 52 Passenger casualty rates by mode:

9 Foreign registered vehicles 53 Reported accidents, vehicles and casualties: by vehicle type and foreign registration Calendar of events affecting road safety and traffic Review topics Accident statistics report form (MG NSRF) Index of topics

10 . General overview and progress towards casualty reduction targets Road Safety Research and Statistics, Department for Transport Summary This article reviews the main trends in the number of reported road accident casualties in Great Britain in 28 compared with recent years. It also reports progress towards the Government s 2 casualty reduction targets for Great Britain. Figures are derived from information about accidents reported to the police (see Article 5). In 28: There were a total of 23,95 reported casualties of all severities, 7 per cent lower than in 27. 2,538 people were killed, 4 per cent lower than in 27, 26,34 were seriously injured (down 6 per cent) and 22,333 were slightly injured (down 7 per cent). The number of fatalities fell for almost all types of road user, with a fall of 2 per cent for car occupants, per cent for pedestrians, 6 per cent for motorcyclists and 5 per cent for pedal cyclists. In 2, the Government set a new target for a reduction in the number of casualties in road accidents. By 2 the aim is to achieve, compared with the average for , a 4 per cent reduction in the number of people killed or seriously injured in accidents reported to the police; a 5 per cent reduction in the number of children killed or seriously injured; and a per cent reduction in the slight casualty rate. Compared with the baseline ( average), in 28: The number of reported killed or seriously injured casualties was 4 per cent lower; The number of children killed or seriously injured was 59 per cent lower; and The slight casualty rate was 36 per cent lower. Overall traffic rose by an estimated 6 per cent. Charts showing progress towards targets and trends in road accident casualties compared with traffic can be found in the Annex to this article. Table a: Reported road accident casualties by severity: GB 28 Number 28 Percentage change over: average average Killed 3,578 3,72 2,946 2, of which children Seriously injured 44,78 28,673 27,774 26, Killed or seriously injured 47,656 3,845 3,72 28, of which children 6,86 3,294 3,9 2, Slightly injured 272, ,559 27,6 22, All severities 39, ,44 247,78 23, Traffic 4,443 5,2 5,72 5,37 6 KSI rate Slight casualty rate Traffic in million vehicle kilometres; rates per million vehicle kilometres. 6

11 Part : Trends in reported road accident casualties Killed or seriously injured (KSI) casualties The Government s main casualty reduction targets relate to the number of road users killed or seriously injured (KSI) in accidents reported to the police. Overall, the reported number of people killed or seriously injured fell by 7 per cent between 27 and 28 and by a total of 4 per cent from the average. The fall in KSI casualties has occurred despite a rise in overall traffic levels of around 6 per cent between the baseline and 28. Between 27 and 28 traffic fell by per cent. Compared with the baseline, there have been reductions in the number of reported KSI casualties (of between 3 and 5 per cent) for all types of road user, with the exception of motorcyclists where the number fell by 7 per cent. Over this period motorcycle traffic increased by 33 per cent in total (more than any other road user type), so that the KSI casualty rate for motorcyclists fell by 3 per cent Around 2 out of every 5 people killed or seriously injured are car occupants. Car occupant KSI casualties fell 49 per cent from the baseline. Over the same period car traffic increased by 2 per cent. Table b: Reported killed or seriously injured casualties by road user type: GB 28 Number 28 Percentage change over: average average (traffic) Pedestrians Pedal cyclists Motorcycle users Car users Bus/coach users Other road users,669 3,732 6,475 23,254 76,8 7,5 2,442 6,484 4, ,88 6,924 2,564 6,737 2, ,73 6,642 2,565 6,49, All road users 47,656 3,845 3,72 28, Chart a: Reported killed or seriously injured casualties by road user type: Detailed information on trends in traffic in Great Britain over the last decade can be found in the Department s annual bulletin: 7

12 KSI casulaties by region The map shows, for each Government Office region, the number of reported killed or seriously injured casualties for the baseline ( average) and 28. In 28, the South East and London regions had the greatest number of KSI casualties (together accounting for more than a quarter of the total), reflecting the larger populations in these areas. The biggest percentage reductions in KSI casualties were in the West Midlands (53 per cent) and London (47 per cent) with a fall of at least 2 per cent in every region. The level of reduction in KSI casualties will be affected by differing trends in traffic and variations in the type of road user involved in accidents across regions. Map a: Reported killed or seriously injured casualties by Government Office Region: average and 28 8

13 Child KSI casualties The Government has set a separate target to reduce the number of children reported killed or seriously injured by 5 per cent from the baseline. In 28, the number of children killed or seriously injured was 2,87-59 per cent below the baseline and 9 per cent lower than in 27. Compared with the baseline, the number of reported child KSI casualties more than halved by 28 for pedestrians, pedal cyclists and car users. The majority of child KSI casualties are pedestrians, accounting for 57 per cent of the total in 28. Compared with 27, there was a 6 per cent fall in child pedestrian KSI casualties, a 7 per cent fall in car occupant KSI casualties but a fall of 2 per cent in child pedal cyclist KSI casualties. In 28, around 2 of every 3 child KSI casualties were male. The number of children aged 2 5 killed or seriously injured has fallen slightly less than other child age groups, by around 52 per cent since the baseline. Table c: Children reported killed or seriously injured by road user type: GB 28 Number 28 Percentage change over: average average Pedestrians Pedal cyclists Car users Other road users 4,67,29, , , , Males Females 4,42 2,457 2,7,87 2,7,83, Age 4 Age 5 8 Age 9 Age ,657,592 2, , , , All children 6,86 3,294 3,9 2, Chart b: Children reported killed or seriously injured by road user type:

14 Fatalities There were a total of 2,538 fatalities in road accidents in 28, 48 fewer than in 27 but still an average of 7 deaths per day. Chart c shows the trends in reported fatal, serious and slight casualties. Trends in fatalities and serious injuries were similar between 99 and 998, with a divergence between 998 and 25; deaths falling by 6 per cent and serious injuries by 29 per cent. However, between 25 and 28, the number of deaths fell by 2 per cent, compared with a per cent fall in serious injuries. Car occupants, pedestrians and motorcyclists account for the vast majority of deaths. In 28, pedestrian fatalities were 43 per cent below the baseline and car occupant fatalities 29 per cent below the baseline, but the number of motorcycle deaths was 6 per cent higher. However, when adjusting for changes in traffic, fatality rates for all road users - including motorcyclists - have fallen from the baseline. Between 27 and 28 fatalities fell by at least per cent for all main road user types. The number of children killed in reported road accidents has fallen by considerably more than the overall fatalities figure, by 52 per cent from the baseline. Although the number of child fatalities increased slightly to 24 in 28 from 2 in 27, this was still well below the 26 figure (69). Table d: Reported fatalities by road user type: GB 28 Number 28 Percentage change over: average average (traffic) Pedestrians, Pedal cyclists Motorcycle users Car users,762,62,432, Bus/coach users Other road users All road users 3,578 3,72 2,946 2, of which children Chart c: Indexed reported casualties by severity: 99 28

15 The 4 per cent reduction in deaths between 27 and 28 is the largest percentage fall in a single year in the post war period, and follows a 7 per cent fall between 26 and 27. Chart d shows the rolling year percentage reduction by month over the last two years; this shows how the reductions began in mid-27 and have been sustained throughout 28. Chart d: Reported road fatalities in Great Britain: Rolling 2 month percentage reduction Although motor vehicle traffic fell by per cent between 27 and 28 (with a fall of 8 per cent for motorcyclists), this is not sufficient to explain the size of the reduction in deaths over this period, as can been seen from charts e and f which show fatality rates per billion vehicle kilometres for different road user groups: In 28 there were 3. car occupants killed per billion vehicle kilometres travelled. This rate has fallen sharply in the last two years, by 22 per cent from 26, compared to a 2 per cent fall in the previous ten years. The pedestrian fatality rate (pedestrian deaths compared with total traffic) has fallen steadily in recent years. In 28 it was 5 per cent below the average and per cent lower than in 27. Chart e: Fatality rates by vehicle type (pedestrians vs. all traffic): GB

16 Motorcyclists have the highest fatality rate of any road user group. In 28, 96 motorcyclists were killed per billion kilometres travelled by motorcycles. However, this is 9 per cent lower than in 27 and 2 per cent below the average. Having remained fairly steady between 24 and 27 the pedal cycle fatality rate fell 24 per cent from 27 to 28, and was 47 per cent below the average. Chart f: Motorcyclist and pedal cyclist fatality rates: GB There are many possible reasons for the large reduction in fatalities, and further analysis is required to understand this (as far as is possible). However, analysis presented elsewhere in this publication provides indications of some key trends: Part 2 of this article looks in more detail at individual road user groups. The number of deaths in accidents involving young car drivers (aged 7 24) fell by 22 per cent, and almost half the overall drop in road deaths was in accidents involving a young car driver. The biggest reduction in motorcyclist fatalities was among those riding bigger bikes (over 25cc) on non-built up roads. Article 3 looks at drinking and driving. This shows that the number of people killed in drink-drive accidents fell considerably from 56 in 26 to 4 in 27, with a provisional figure of 43 for 28 (7 per cent of all deaths). This suggests that a reduction in fatal drink-drive accidents contributed to the overall reduction in fatalities between 26 and 27, but not between 27 and 28 Article 4 contains details of contributory factors for fatal accidents. The patterns shown are broadly similar to those seen in previous years. The tables section in this publication contains a number of tables showing time series of fatalities (for example, tables 3 6 and 8 ). 2

17 Slightly injured casualties In addition to targets for killed and seriously injured casualties, it is the aim to reduce the rate of reported slight casualties by per cent by 2, compared with the baseline. In 28, this rate was 39 slight casualties per million vehicle kilometres, 36 per cent below the baseline level. Compared with the baseline, the biggest reductions in the reported slight casualty rates have been for pedestrians, with falls of at least 34 per cent for each of the main road user types. Between 27 and 28 the number and rate of slight casualties fell for all road user types except Bus and Coach users, with an 8 per cent fall for motorcyclists. Whilst the majority (over two thirds) of slight casualties are car occupants, the highest rates (per million vehicle kilometres) are for motorcycle users, followed by pedal cyclists. Table e: Reported slightly injured casualties by road user type: GB 28 Number 28 Percentage change over: average average Pedestrians Rate 34, , , , Pedal cyclists Rate 2, , , , Motorcycle users Rate 7, , , , Car users Rate 8, , , , Bus/coach users Rate 8, , , , All road users 2 Rate 272, , , , Slight casualty rates for pedestrians are calculated using total vehicle kilometres for all vehicles 2 Total includes other road users Chart g: Rate of reported slightly injured casualties per million vehicle km: GB

18 Part 2: Reported casualties by road user type This section provides the main figures and some analysis for each of the main groups of road user. Chart h shows the proportion of each road user type for the three different severities of casualty in 28: Car occupants were the largest group for all severities, accounting for 68 per cent of reported slight casualties and 5 per cent of fatalities. Pedestrians accounted for 23 Chart h: Proportion of reported casualties by road user type per cent of reported deaths and severity: GB 28 and serious injuries but only per cent of slight injuries. Similarly, 9 per cent of all fatalities were motorcycle users, but only 8 per cent of those slightly injured. Together, car occupants, pedestrians and motorcyclists accounted for 9 per cent of deaths, and 86 per cent of all reported casualties. Of the remainder, pedal cyclists made up 7 per cent and bus or coach users 3 per cent of all casualties. Overall, around 7 of every people reported killed or seriously injured in road accidents were male, but again this varies by road user type - in 28, 9 out of motorcyclist and more than 8 out of pedal cyclist KSI casualties were men, compared with around 6 in pedestrians and car occupants. Chart i shows how the number killed or seriously injured varies by age and road user. The overall number of KSI casualties is highest for ages 7 and 8. For ages 2 to 5 and from age 8 onwards, most KSI casualties are pedestrians. Between the ages of 6 and 6, most casualties are car or motorcycle users. Detailed figures relating to the number of reported road accident casualties by age, gender and road user type can be found in the tables section. Chart i: Reported KSI casualties by road user type and age: GB 28 4

19 Pedestrian casualties Total reported pedestrian casualties have decreased by 6 per cent from 3,9 in 27 to 28,482 in 28, and were 39 per cent below the baseline average. Overall pedestrian fatalities fell by per cent from 27 to 28, although this varied by age group. Child pedestrian fatalities were unchanged in 28 at 57 deaths, 57 per cent below the average baseline. Ten per cent of all pedestrian fatalities were children (aged 5 years old); however this proportion rose to 3 per cent for all pedestrian casualties. The number of adult pedestrians killed aged 6 to 59 years old fell by per cent, from 34 in 27 to 272 in 28. There was a 4 per cent decrease in the number of pedestrian fatalities aged 6 years old and over, from 28 in 27 to 243 in 28. Adults 6 years old and over accounted for 42 per cent of all pedestrian fatalities but only 5 per cent of all casualties. The rate of reported pedestrian casualties per, population has been falling and in 28 was 42 per cent lower than the baseline and 6 per cent lower than in 27. The rate for pedestrian casualties aged 6 years old and over was the lowest of all age groups, with child pedestrian casualties being the highest (32 pedestrian casualties per, population for 6 year olds and over, compared to 78 for 5 year olds). Table f: Reported pedestrian casualties by age: GB 28 Number 28 Percentage change over: average average Children ( 5) Killed Serious 4,34,954,842, Slight 4,382 8,6 7,628 6, All 8,548,3 9,527 8, Adults (6 59) Killed Serious 4,38 3,2 3,93 3,3 3 3 Slight 5,6 2,6,965, All 9,732 5,55 5,362 4, Adults (6+) Killed Serious 2,42,7,222,26 44 Slight 4,49 2,82 2,8 2, All 7,4 4,259 4,34 4,8 3 4 All Killed, Serious,662 6,376 6,278 6, Slight 34,874 23,93 23,267 2, All 46,543 3,982 3,9 28, Casualty rate per, population KSI Slight All Includes cases where age not reported. Tables 3 34 in the tables section provide a further breakdown of pedestrian casualties. 5

20 Pedal cycle casualties Overall reported pedal cycle casualties in 28 remained at a similar level to 27, but have decreased by 33 per cent from the baseline. The number of pedal cycle fatalities fell by 5 per cent from 36 in 27 to 5 in 28, a 38 per cent decrease from the average baseline. However, the number of reported seriously injured pedal cyclists increased by per cent from 2,428 in 27 to 2,45 in 28. Table g: Reported pedal cyclist casualties: GB 28 Number 28 Percentage change over: average average Killed Serious 3,546 2,296 2,428 2,45 3 Slight 2,653 3,754 3,63 3, Total 24,385 6,96 6,95 6, Pedal cycle Traffic Casualty Rate 2 KSI Slight All million vehicle kilometres. 2 Rate per million vehicle kilometres. Pedal cycle traffic levels have fluctuated in recent years, but the trend has been generally upward. Chart j shows that trends in pedal cyclists killed and injured have followed broadly similar trends since 994, with levelling off in casualties from around 22 coinciding with increasing pedal cycle traffic. Chart j: Pedal cycle traffic and reported casualties by severity: GB

21 8 per cent of reported pedal cycle casualties were male, as were 84 per cent of pedal cycle fatalities. 58 per cent of all pedal cycle casualties were 6 59 year old male pedal cyclists, compared to 57 per cent for pedal cycle fatalities. Just over a fifth of pedal cycle casualties were children ( 5 years old). However only per cent of pedal cycle fatalities were children. The number of reported child pedal cycle casualties has fallen by 58 per cent since the average baseline, from 7,85 in the baseline to 3,36 in 28. The number of female child casualties has fallen more than for male casualties (65 per cent compared to 56 per cent reduction) Tables 29a, b and c in the tables section analyse reported casualties by severity, day, road user type and hour of day. Fifty one per cent of pedal cycle casualties occurred during the hours of 7am am and 4pm 7pm. This proportion was slightly higher for accidents on Monday to Thursday (57 per cent) and lowest on Sundays (32 per cent), and is likely to be related to school and work travel. The proportions are similar for both child and adult casualties. Chart k looks at the number of reported killed or seriously injured pedal cycle casualties, by month and age of casualty for the past 5 years. In recent years, the number of adult (6 years old and over) pedal cyclists killed or seriously injured tended to peak in June and July. For children (aged to 5 years old), the peak in the number of pedal cycle KSI casualties tends to coincide with summer holidays. Chart k: Reported pedal cycle KSI casualties, by month, and age of casualty: GB

22 Motorcycle user casualties Reported motorcycle casualties decreased by 8 per cent from 23,459 in 27 to 2,55 in 28, and were per cent lower than the average baseline. Motorcycle traffic fell by 8 per cent compared to 27, but was 33 per cent higher than the average baseline. Motorcycle fatalities fell by 6 per cent from 588 in 27 to 493 in 28. However, since the average baseline, motorcycle fatalities have increased by 6 per cent. There was a per cent fall in the number of reported serious motorcycle casualties, resulting in a per cent decrease in the number of KSI motorcycle casualties, from 6,737, in 27 to 6,49 in 28. Due to the fall in motorcycle traffic, the motorcycle casualty rate fell very slightly from 42 motorcycle casualties per million vehicle kilometres in 27 to 49 in 28. The motorcycle casualty rate rose for slight casualties, but fell for both serious injuries and fatalities. Table h: Reported motorcycle user casualties: GB 28 Number 28 Percentage change over: average average Killed Serious 6,8 5,885 6,49 5,556 8 Slight 7,547 6,842 6,722 5,5 7 2 Total 24,23 23,326 23,459 2,55 8 Motorcycle Traffic Casualty Rate 2 KSI Slight All million vehicle kilometres. 2 Rate per million vehicle kilometres. 65 per cent of motorcycle fatalities occurred in rural areas, compared to 47 per cent for serious motorcycle casualties and 3 per cent for slight motorcycle casualties. 4 per cent of riders of motorcycles less than 5cc involved in personal injury road accidents were aged 6 years. A further 9 per cent were 7 years old. This is in contrast to motorcycles greater than 5cc, where 57 per cent of riders were aged 3 49 years. Chart l shows the trends in reported motorcycle casualties per million vehicle kilometres and motorcycle traffic, indexed to the average. Motorcycle traffic increased from the average baseline till 23. Since 23, the traffic has been fairly volatile, with the 28 traffic figure being at a similar level to the 26 figure, 33 per cent greater than the average baseline. Motorcycle casualty rates for all severities have declined over the same period. Motorcycle fatalities per million vehicle kilometres have shown a lower decrease since the compared to injuries. 8

23 Chart l: Reported motorcycle casualties per million vehicle kilometres, and traffic: GB Chart m shows the number of reported motorcyclists killed by road type and engine size in the past years. It shows that most of the fall in fatalities in 28 came amongst riders of motorcycles over 25cc on non built-up roads, falling from 324 in 27 to 24 in 28, a 26 per cent drop. 7 per cent of motorcycle fatalities were riding motorcycles greater than 5cc. In 28, 347 motorcycle fatalities were on these vehicles, compared to 443 in 27; a 22 per cent decrease. There has been a rise in the number of fatalities for riders of motorcycles with an engine capacity under 25cc - fatalities have increased from 74 in 27 to 79 in 28. Chart m: Reported motorcyclist fatalities by road type and engine size: GB

24 Car occupant casualties Reported car occupant casualties, given in table i, were 8 per cent lower than in 27, falling from 6,433 in 27 to 49,88 in 28. The 28 figure reflects a 27 per cent decrease since the average baseline figure. Car occupant fatalities decreased by 2 per cent from 27, with falls for both car drivers and passengers (9 per cent and 9 per cent respectively). Compared to the average car driver deaths have fallen more slowly than for passengers, 24 per cent lower than the baseline compared to 38 per cent for car passengers. Car traffic has increased by 2 per cent since the average baseline, but is per cent lower than the 27 level. The number of reported killed or seriously injured car occupants per million vehicle kilometres has fallen by 7 per cent since 27, and 54 per cent from the average baseline. The slight car casualty rates were 7 per cent and 32 per cent lower respectively over the same time periods. Table i: Reported car user casualties: GB 28 Number 28 Percentage change over: average average Drivers Killed Serious Slight Total,28 3,56 3,324 27,958,66 8,239 5,698 5, ,537,62 9, 86 7,6 92,985, Passengers Killed Serious Slight Total 634 7,985 66,7 75, ,43 5,48 55, ,998 47,845 52, ,65 44,235 48, All Killed Serious Slight Total,762 2,492 8,34 23,288,62 2,642 56,746 7,,432,535 48,466 6,433,257,7 37,22 49, Car traffic 3,585 4,26 4,4 4,7 2 Casualty rate 2 KSI Slight All million vehicle kilometres. 2 Rate per million vehicle kilometres. Chart n shows the number of reported car occupants killed by age group. In 28 there were 46 fatalities amongst car occupants aged This was an 8 per cent fall from 27 and a 28 per cent fall from the 26 figure. However, this figure is just 23 per cent below the average, while for all the other age groups shown fatalities have fallen at least 27 per cent. Child car occupant fatalities rose from 46 in 27 to 49 in 28. However, this is 37 per cent lower than the average. 2

25 Chart n: Reported car occupant fatalities by age group: GB Table 38 in the tables section looks at the age distribution of car drivers involved in personal injury road accidents reported to police, by gender. Chart o shows the number of fatalities resulting from accidents involving at least one young car driver (7 24 years old): Fatalities in reported accidents involving young car drivers accounted for 25 per cent of all road deaths in 28. The number of fatalities in accidents involving young car drivers fell by 22 per cent from 87 in 27 to 635 in 28 a reduction of 82 deaths, out of a total fall in road deaths of 48 between 27 and 28. The number of young car drivers killed decreased by 6 per cent from the average baseline (to 256 in 28), whilst passengers fatalities of young car drivers decreased by 43 per cent (to 4). The number of other participants killed in accidents with a young car driver (occupants of other vehicles and pedestrians in the accident) fell by 44 per cent (to 238). Chart o: Reported fatalities in accidents with young car drivers (age 7 to 24): GB

26 KSI casualties in reported accidents involving young car drivers fell by 2 per cent between 27 and 28 (to 6,855) and accounted for 24 per cent of all KSI casualties in 28. Just over a fifth of all car occupants killed or seriously injured were young car drivers. Killed or seriously injured young car drivers have decreased by 43 per cent (to 2,268) from the average baseline, whilst passengers of young car drivers have decreased by 5 per cent (to,479) and other participants (occupants of other vehicles and pedestrians in the accident) have decreased by 5 per cent (to 3,8). Other road user casualties Reported bus and coach casualties decreased by 2 per cent compared with 27, and were 28 per cent lower in 28 than the baseline average. The number of fatalities fell from 2 in 27 to 6 in 28. The number of serious injuries fell by 4 per cent in 28 from 27, but was 39 per cent lower than the average. Care should be exercised when comparing these percentage changes with other road user types since these numbers are small and are therefore liable to fluctuations. In 28, bus and coach traffic decreased by per cent from 27, but this is still 4 per cent higher than the average baseline. Reported light goods vehicle casualties in 28 were 8 per cent lower than in 27, and 34 per cent lower than the average. Light goods traffic has remained unchanged since 27, but this is 46 per cent higher than the average baseline. The casualty rate has decreased much more, 8 per cent since 27 and 55 per cent since the baseline. Deaths among light goods vehicle users however fell by 22 per cent, from 58 in 27 to 43 in 28. This represents a 34 per cent decrease compared to the average baseline. Reported heavy goods vehicle occupant casualties have decreased by 22 per cent from 27 and 42 per cent compared with the average baseline. Fatalities fell by 56 per cent, from 52 in 27 to 23 in 28. Heavy goods vehicle traffic has decreased by 2 per cent from 27, but is still per cent higher than the average baseline, resulting in a reduction of 2 per cent and 47 per cent respectively in the overall casualty rate for heavy goods vehicle occupants. 22

27 Table j: Reported other road user casualties: GB 28 Number 28 Percentage change over: average average Bus and Coach Killed Serious Slight 8,883 6,827 6,624 6, Total 9,598 7,253 7,79 6, Bus/Coach traffic Light goods vehicle Killed Serious Slight 6,4 5,35 4,846 4, Total 7,424 5,94 5,34 4, Light goods traffic Heavy goods vehicle Killed Serious Slight 2,76 2,47 2,3, Total 3,338 2,53 2,476, Heavy goods traffic million vehicle kilometres. 23

28 Annex: Progress towards targets and long term trends Chart p: Reported killed or seriously injured casualties: Chart q: Reported killed or seriously injured child casualties: Chart r: Rate of reported slightly injured casualties per million vehicle kilometres:

29 Chart s: Indices of population, vehicle stock, motor traffic and reported casualties: Chart t: Indices of population, vehicle stock, motor traffic and reported casualties: average = Population Traffic Injured Vehicles Killed

30 2. A valuation of road accidents and casualties in Great Britain in 28 Shamanthy Ganeshan, Integrated Transport Economics & Appraisal, Department for Transport. Summary This article provides the latest Department for Transport estimates of the values for prevention of road accidents and casualties for use in the appraisal of transport schemes and gives an estimate of the total value of road accidents in Great Britain in 28. Since 993, the valuation of both fatal and non-fatal casualties has been based on a consistent willingness to pay (WTP) approach. This approach encompasses all aspects of the valuation of casualties, including the human costs, which reflect pain, grief, suffering; the direct economic costs of lost output and the medical costs associated with reported road accident injuries. The methodology used to value the cost of casualties was described in an article in Road Accidents Great Britain 994 (Kate McMahon, Road Safety Division, Department for Transport). More detailed information on the valuation of the benefits of preventing accidents and casualties is published by the Department for Transport in Transport Analysis Guidance Unit 3.4., The Accident Sub-Objective. The total value of prevention of all road accidents in 28 was estimated to be 7.9bn. This does not make full allowance for under-reporting of injury accidents, but does include an estimate of the cost of damage only collisions. Casualties The values for the prevention of fatal, serious and slight casualties include the following elements of cost: Loss of output due to injury. This is calculated as the present value of the expected loss of earnings plus non-wage payments made by employers. Ambulance costs and the costs of hospital treatment. The human costs of casualties. These are based on willingness to pay to avoid pain, grief and suffering to the casualty, relatives and friends, as well as intrinsic loss of enjoyment of life in the case of fatalities. Accidents The average value of preventing a fatal accident is greater than the value of preventing a fatality. This applies for each level of severity. This is for two reasons, the first being that an injury accident is classified according to the most severe casualty but will on average involve more than one casualty. For example, in 28 a fatal accident on average involved.8 fatalities,.32 serious casualties and.48 slight casualties. The second reason is that there are some costs which are part of the valuation of an injury accident but which are not specific to casualties. These are: 26

31 Costs of damage to vehicles and property. Police costs and administrative costs of accident insurance. Valuation of the benefits of prevention of accidents Table 2a gives the average values of prevention of reported road accidents and casualties in 28 prices; Table 2b gives the average value of prevention of reported injury accidents by different types of road. Table 2a: Average Value of Prevention per reported Casualty and per reported Accident: GB 28 June 28 Accident/casualty type cost per casualty cost per accident Fatal Serious Slight Average for all Severities Damage only,683,8 89,2 4,6 52,6,96,2 28, 22,6 75, 2, Table 2b: Average value of prevention of reported road accidents by road type: GB 28 June 28 Accident Type Built-up roads Non Built-up roads Motorways All Roads Fatal Serious Slight All injury Damage only,86,2 29,8 2,5 59,7,9,973,6 234,5 25,2 2, 2,8 2,64,5 243,5 29,9 89, 2,7,96,2 28, 22,6 75, 2, The total value of prevention of road accidents in GB in 28 Estimates of the total value of prevention of road casualties and road accidents in Great Britain during 28 are provided below. The estimates were derived using the values for prevention of casualties and accidents listed above, and are cost benefit values that represent the benefits which would be obtained by prevention of road accidents. The estimates do not represent actual costs incurred as the result of reported road accidents. A total of 2,34 fatal accidents, 23,2 serious accidents and 45,29 slight accidents were reported in 28. In cost-benefit terms the value of prevention of these 7,59 reported injury accidents is estimated to have been 2,79m in 28 prices and values. No estimate has yet been made of the non-medical costs related to injury accidents not reported to the police. However, there were an estimated 2.6 million damage-only accidents valued at a further 5,3m. A proportion of these may have involved injuries not reported to the police. The total value of prevention of reported injury and damage only road accidents in 28 was therefore estimated to have been 7,92m. This estimate relates to the total value to the community of the benefits of prevention of road accidents. Some costs, such as lost output, will not be borne exclusively by casualties themselves, since the taxation and social security systems will ensure that the burden of lost output will be shared by the population at large. Whereas some elements of cost, e.g. property damage, represent direct costs that will be incurred as the result of road accidents, others like human costs represent the benefit of avoidance of risk of a reported road accident, rather than values of the consequences of an accident. The tables below give the total value of prevention of road accidents by severity and element of cost (Table 2c), and by severity and category of road (Table 2d), without attempting to allocate costs by responsibility or final incidence. 27

32 Table 2c: Total value of prevention of reported accidents by severity and element of cost: GB 28 Cost Element m June 28 Casualty related costs Accident related costs Accident Severity Lost Medical and Human Police Insurance Damage to Total output Ambulance costs cost and admin property Fatal,49 2, ,5 Serious , , Slight , ,3 All injury 2, , ,8 Damage only (see text) 4 4,98 5, All accidents 2, , ,58 7,9 Table 2d: Total value of prevention per reporeted accident by severity and class of road: GB 28 m June 28 Accident severity Built-up roads Non Built-up roads Motorway All roads Fatal Serious Slight All injury Damage only All accidents,84 3,28 2,3 7,43 4,4,57 2,35, , , ,5 5, 3,3 2,8 5, 7,9 During 28, 85 per cent of reported accidents occurred on built-up roads, but these accounted for only 65 per cent of the total value of injury accidents, because they were, on average, less severe than on other roads, having fewer casualties per accident and a lower proportion of fatal and serious injuries. Non built-up roads accounted for 3 per cent of report accidents and 3 per cent of value, and 2 per cent of accidents with 4 per cent of value occurred on motorways. The lesser severity of accidents on built-up roads is shown in Table 2b, where the average value of prevention per accident on built-up roads is less than half the average value on non built-up roads. Further information Further information on the method used to derive the values of preventing road accidents and casualties, together with guidance on how to apply them can be found in Transport Analysis Guidance Unit 3.4., The Accident Sub-Objective, which is available at: In the event that additional information is required, please contact a member of the Integrated Transport Economics and Appraisal division by telephone on or via itea@dft.gov.uk. *The figures in this article are outside the scope of National Statistics. 28

33 3. Drinking and driving Yingbo Xu, Road Safety Research and Statistics, Department for Transport Summary This article presents updated statistics on reported drinking and driving accidents and casualties. It first explains how drink drive accidents and casualties are defined, and the alcohol test limits that apply in Great Britain. The article then looks at an analysis of the characteristics of drink drive accidents and casualties. A description of the sources of data used to produce the drink drive estimates, and a discussion of their reliability are available in the Annex. In 28, it was estimated that 3,2 reported casualties (6 per cent of all road casualties) occurred when someone was driving whilst over the legal alcohol limit. The provisional number of people estimated to have been killed in drink drive accidents was 43 in 28 (7 per cent of all road fatalities), an increase of 2 fatalities compared to 27. The provisional number of KSI (killed or seriously injured) casualties in 28 was 2,6, less than a quarter of the 98 level and 5 per cent below the 27 level. Provisional figures for the number of slight casualties in 28 fell 7 per cent since 27, from,85 to,97. Reported drink drive accident limits and definitions For the purposes of these drink drive statistics, a reported drink drive accident is defined as being a collision on a public road reported to police in which someone is killed or injured and where one or more of the motor vehicle drivers or riders involved either refused to give a breath test specimen when requested to do so by the police (other than when incapable of doing so for medical reasons), or one of the following: i) failed a roadside breath test by registering over 35 micrograms of alcohol per millilitres of breath ii) died and was subsequently found to have more than 8 milligrams of alcohol per millilitres of blood. Please note that where reference is made to drivers/riders over the legal limit this includes those who refused a breath test as well as those failing a test. Drink drive casualties are defined as all road users killed or injured in a drink drive accident. However, not all drink drive accidents are detected in this way, as there are some drivers involved for whom neither of the above test results are available, even though they were over the legal limit. The Department s statistics therefore are adjusted to allow for this in order to produce a better estimate of the number of drink drive accidents and casualties. The reasons for the unavailability of some data, the methods of adjustment and the main data sources used are described in more detail in the Annex. Estimates for 28 are provisional. This is due to Coroners data being available for analysis around eighteen months in arrears. Around 57 per cent expected to be available ultimately were available for inclusion in the provisional estimates in this article. For this reason, the detailed analysis in this article is based on 27 data. Further information about the nature of the provisional estimates is available in the Annex. 29

34 Analysis of reported drink drive data Table 3a shows estimates of the number of reported drink drive accidents and resulting casualties in Great Britain for 979 to 28. Table 3a: Estimated number of reported drink drive accidents and casualties: GB Number Accidents Casualties Year Fatal Serious Slight Total Killed Serious Slight Total 979,38 5,63 2,46 9,47,64 8,3 2,49 3,43 98,28 5,43,86 8,57,45 7,97 2,42 29,83 98,2 4,94,9 7,4,42 7,37 9,6 27,95 982,3 5,42 2,7 8,8,55 8, 2,66 3, ,75,43 7,3, 6,8 8,6 26,52 984, 4,79,54 7,32,7 6,82 9,4 27, ,9,46 7,26,4 6,8 9,38 27, ,59,5 6, ,44 9,22 26, ,22,56 5,56 9 5,9 7,67 24, ,66,9 4, , 6,86 22, ,39,3 4,39 8 4,79 6,62 22, ,9 9,65 3,2 76 4,9 5,55 2, ,59 8,53, ,6 3,6 7, ,36 7,89, ,28 2,77 6, ,87 7,6 9, ,66,78 4, ,9 7,33 9,9 54 2,84,78 5, ,4 7,59,8 54 3, 2,45 6, ,5 8,24, , 3,45 7, ,4 8,,7 55 2,94 3,3 6, ,86 7,84, 46 2,52 2,6 5, ,85 8,8,5 46 2,47 3,98 6,9 2 45,95 9,4,8 53 2,54 4,99 8, ,2 9,78 2, ,7 5,55 8, ,5,62 3,5 55 2,79 6,76 2, 23 5,97 9,93 2,4 58 2,59 5,82 8, ,79 8,9,2 58 2,34 4,6 6, ,54 8,6,7 55 2,9 2,76 5, ,48 7,43 9,4 56,97,84 4, ,4 7,52 9,28 4,76,85 4,2 28 P 38,28 6,98 8,64 43,63,97 3,2 P Provisional data. The sample of fatality data from Coroners for 27 has now been finalised but 28 estimates are based on a reduced sample of coroners returns and may be biassed. They remain provisional until more complete information for 28 is available. Provisional figures in 28 show there were 8,64 reported personal injury road accidents involving at least one driver/rider over the legal alcohol limit, of which 38 were fatal accidents. This represents a 7 per cent decrease in all drink drive accidents since 27, but a 3 per cent increase in fatal accidents. Serious accidents fell to a low of,28, whilst slight accidents fell to 6,98. In 28, there were 3,2 casualties resulting from drink drive accidents, a 7 per cent decrease since 27. The provisional number of fatalities rose to 43 in 28, an increase of 5 per cent from 27. The number of drink drive fatalities accounts for 7 per cent of all road accident fatalities. The number of seriously injured drink drive casualties has been declining gradually since 22. The provisional figure of,63 in 28 was the lowest since the series began, and represents a 7 per cent decrease from 27. (Chart 3a) Slight casualties fell 7 per cent from 27, from,85 to,97 in 28. 3

35 Chart 3a: Estimated number of killed or seriously injured reported drink drive casualties, GB Characteristics of reported drink drive casualties Table 3b shows the percentage of driver and rider fatalities in reported accidents who were over the legal alcohol limit by age group during 998 to 28. The proportion had fallen considerably since the early 98s, when around a third of drivers and riders killed were over the limit. It has since remained at about one in five (dipping to one in six between 997 and 999). Provisional figures for 28 indicate a rise in the percentage of car and other motor vehicle driver fatalities who were over the limit for all age groups, whilst motorcycle riders showed an overall decrease. Table 3b: Drivers and riders killed in reported accidents: Percentage over the legal blood alcohol limit: GB Motorcycle riders Cars and other motor vehicles All Year Age 6 9 Age 2 29 Age 3 39 Age 4+ All Ages Age 6 9 Age 2 29 Age 3 39 Age 4+ All Ages P () (2) Source: Coroners and Procurators Fiscal only P Provisional data. The sample size for 28 is not yet sufficient to give a full age breakdown. Age groups Age groups 3+ 3

36 Women are much less likely to be involved in reported drink drive accidents as drivers than men. However, Table 3c shows that nearly a third of the total casualties in drink drive accidents were women. It is estimated that in 27 there were around 46 pedestrian casualties and 2 pedal cyclist casualties in accidents with a driver over the legal alcohol limit. Table 3c: Estimated number of reported drink drive casualties, by casualty type: GB 27 Number Car drivers Motor- Car Pedestrians Cyclists cyclists Over limit Under limit passenger Other Male Female Total Killed or seriously injured casualties , All ages , ,7 Total casualties ,5 49 2,2 4 3,79,54 5, ,9,73, ,92 2,2 7, All ages ,4 2,5 4,48 6 9,59 4,43 4,2 Includes age not recorded. Chart 3b shows the percentage of killed drivers/riders within each blood alcohol content (BAC) category, by age. People aged 5 59 years old, and 6 years old and over had the highest proportion of killed drivers/riders with no alcohol present in their blood (73 and 8 per cent respectively). Conversely, 2 24 year olds had the lowest proportion of killed drivers with no alcohol present, but the highest for killed drivers/riders over the legal alcohol limit (29 per cent for those aged 2 24 compared to 4 per cent for 6+ year olds). Drivers/riders killed who were in the 2 24 year old age group also had the highest proportion for blood levels over twice the legal alcohol limit. 32

37 Chart 3b: Proportion of all killed drivers resulting from reported accidents in each BAC category, by age: GB 27 Source: Coroners and Procurators Fiscal only Chart 3c shows the number of killed or seriously injured (KSI) casualties resulting from personal injury road accidents where a young driver/rider (7 24 years old) was over the legal alcohol limit. Chart 3c: Killed and seriously injured casualties in reported accidents involving young drivers/riders (7 24 years old) over the legal alcohol limit: GB The number of KSI young driver/rider casualties over the legal alcohol limit was at its highest in 996 and was relatively constant between 998 and 26. However, between 26 and 27 there was a fall of per cent from 3 to 269 (this compares with a 7 per cent fall in total KSI casualties of young drivers/riders aged 7 24). The number of passengers of these young drivers/riders over the legal alcohol limit also showed a peak in 996, but has since declined to around 26 in 27, a fall of per cent compared to 26 (overall KSI casualties of passengers of young driver/riders fell by 7 per cent between 26 and 27). 33

38 The numbers of other participants (e.g. pedestrians and other drivers, either over the age of 24 or not over the legal alcohol limit) declined to 8 in 27, a 6 per cent decline compared to 26. Table 3d is based on 27 Coroners and Procurators Fiscal data using a sample which accounts for around 57 per cent of all reported fatalities in that year. For these fatalities the table shows the percentages exceeding varying levels of blood alcohol for different classes of road user, and the proportion of fatalities exceeding 8mg/ml by time of day. For example, for motorcycle riders, 22 per cent of motorcycle riders killed had more than 9mg of alcohol per ml of blood, whilst per cent had over 8mg/ml (i.e. over the drink drive limit). 2 per cent of motorcycle riders killed had over 2mg/ml. The pedestrian, passenger and cyclist fatalities shown in the table were not necessarily involved in drink drive accidents, as defined earlier in this article, which involve a motor vehicle driver or rider who was over the limit. Also, blood alcohol levels were available for 75 per cent of motorcycle riders but for only 44 per cent of all pedestrian fatalities. The figures may therefore overestimate the proportion of pedestrian fatalities which are over the legal limit since a pedestrian fatality is more likely to be tested if there is a suspicion of alcohol use. In 27, Approximately one in five drivers killed, excluding motorcycles, was over the legal limit for driving a motor vehicle. The rate for motorcycle riders killed was about half the rate for other drivers. Approximately half of the drivers killed between pm and 4 am were over the limit. Ninety two per cent of pedestrians killed between pm and 4 am were over the legal limit for drivers. Table 3d: Blood alcohol levels of reported fatalities aged 6 and over: GB 27 Percentage Cumulative percentage over blood alcohol levels (mg/ml) Sample size Percentage over 8mg/ml time of accident 22: 3:59 4: 2:59 Motorcycle riders Other vehicle drivers Passengers Pedestrians Cyclists Source: Coroners and Procurators Fiscal only 34

39 Characteristics of reported drink drive accidents Table 3e shows that in both 997 and 27 those car drivers aged under thirty had the most drink drive accidents. Young car drivers (aged 7 24) had more drink drive accidents per thousand licence holders and per million miles driven than any other age group. Car drivers aged 6 years old and over had the least. In most age groups, there was a reduction from in both the numbers and rates of drink drive accidents. In contrast, the rates for drivers aged up to 24 have shown little change. Table 3e: Car drivers in reported drink drive road injury accidents: accidents per licence holder and per mile driven, GB 997 and 27 Drink drive accidents per Drink drive accidents per Car driver drink drive accidents thousand licence holders million miles driven Number Under , ,, ,67, , , ,27, or over All ages 3 9,44 8, Based on NTS average 2 Figures based on a small NTS sample. 3 Includes age not known. Sources: National Travel Survey and STATS9 Drinking and driving is a year-round problem. Although the exact pattern varies year on year, the first few months of the year generally have lower numbers of drink drive accidents and casualties than other months of the year. In 27, there is peak in both the number of accidents and casualties in May (Chart 3d). Chart 3d: Estimated number of reported drink drive accidents and casualties, by month: GB 27 35

40 In 27, 63 per cent of all drink drive accidents occurred during Friday, Saturday or Sunday, with about half of these happening during the hours of 9pm to 3am. Chart 3e shows the proportion of drink accidents by time of day in 997 and 27. The proportion of drink drive accidents in the evening in 27 have reduced from the proportion in 997, with particular decreases from 9pm to midnight. Chart 3e: Reported drink drive accidents, by time of day: GB 997 & 27 In 27, 4 per cent of reported drink drive accidents were single vehicle accidents involving no pedestrians. In these accidents there was therefore only one driver/rider over the legal alcohol limit. Forty three per cent of drink drive accidents involved two vehicles, whilst 3 per cent involved more than two vehicles. 36

41 Breath testing Breath testing rates at reported personal injury road accidents fell marginally to 55 per cent in 28. The proportion of drivers and riders failing breath tests had remained relatively consistent at about 4 per cent though this has fallen over the last few years to 3.4 per cent in 28, whilst the percentage of all drivers and riders involved in injury accidents required to take a breath test and subsequently failed has remained at close to 2 per cent throughout the past ten years (Table 3f). Table 3f: Drivers and riders in reported injury road accidents: breath tests and failures: GB Number/percentage a. Total involved 46,4 48,23 399,883 39, ,98 362,33 348,798 33,55 38,28 294,455 b. Total tests requested 24,75 22,7 2,722 96,232 87,276 83,972 83,29 79,27 79,558 62,969 c. Total failed 7,523 7,967 8,96 8,4 8,5 7,427 7,5 6,594 6,278 5,52 Testing rate (b/a x ) Test failure rate (c/b x ) Total failure rate (c/a x ) Overall, 2.7 per cent of men involved in an accident failed a breath test, well over twice the rate for women (.2 per cent). The rate of failure of breath tests generally declined with age, although those aged 7 9 had slightly lower rates than the high proportion recorded by a few (illegal) drivers aged less than 7. (Table 3g) Table 3g: Car drivers in reported personal injury road accidents: breath tests and failures: GB 28 Number/percentage Men Women a: Involved in accident b: Tested c: Failed b as % of a c as % of a a: Involved in accident b: Tested c: Failed b as % of a c as % of a, , 7,769 5,44 3,47 3,52 24,37 5,329 9,94 7, ,76 2,9,83 8,36 8,543 5,346 9,92 6,476 4, ,592,36 9,864 8,453 9,5 6,323 9,2 4,44 2, ,368 6,94 5,878 4,7 5,233 9,57 5,465 2,7, All ages 4,23 85,437 3, ,963 46,97, Includes age not known 37

42 Chart 3f shows that the number of roadside screening breath tests carried out by police has declined in recent years, from about 8 thousand a year in the late 99s to about 6 thousand in the most recent three years. Of these, only about a third were involved in an injury accident. The proportion of tests failed increased from 2 per cent in 999 to 2 per cent in 23, but since then has declined gradually to 6 per cent in 27. The number of convictions (findings of guilt at courts for driving after consuming alcohol or taking drugs) fell from approximately 93 thousand in 26 to around 89 thousand in 27 (see report released by Ministry of Justice at criminalannual.htm) Chart 3f: Reported roadside screening breath tests and breath test failures, England and Wales Source: Home Office 38

43 Annex Blood and breath testing powers The blood alcohol limit became a legal requirement and roadside breath tests were introduced in 967. Evidential breath testing was introduced in 983 to supplement the taking of blood samples. Section 6 of the Road Traffic Act (988) allows the police to test any driver involved in an accident, whether or not anyone is injured. The act also stipulates that, where there has not been a road accident, the police can only take a roadside breath test following a moving traffic offence, or if there is suspicion of alcohol use. A high breath testing rate is acknowledged to have a deterrent effect upon potential drink drivers, although research shows that a lower number of carefully targeted breath tests, which lessen the burden on police resources, can identify a large proportion of drink drivers. In April 996 the Association of Chief Police Officers in England and Wales (ACPO) adopted a policy of breath testing all drivers involved in road accidents which the police deal with or attend, whether injuries are involved or not. Before this, all Scottish police forces, and some in England and Wales, already operated similar policies, but in some cases for injury accidents only. However, not all drivers involved in injury road accidents are breath-tested; either because the police do not attend the accident, or because a driver leaves the scene before a test can be taken or because they are too seriously injured to take a test. Roadside breath testing rates after injury accidents can still vary widely among police forces. Data sources Two sources of data are used to assess the extent and characteristics of drink drive accidents in Great Britain and a third source provides information on compliance with drink drive restrictions. These sources are: i) Coroners data: Information about the level of alcohol in the blood of road accident fatalities aged 6 or over who die within 2 hours of a road accident is provided by Coroners in England and Wales and by Procurators Fiscal in Scotland. ii) STATS9 breath test data: The personal injury road accident reporting system (STATS9) provides data on injury accidents in which the driver or rider survived and was also breath tested at the roadside. If the driver or rider refused to provide a breath test specimen, then they are considered to have failed the test unless they are deemed unable to take the test for medical reasons. (iii) In addition, police force roadside screening breath test data: Information from breath tests carried out at the roadside following a moving traffic offence, road accident or suspicion of alcohol use is available for England and Wales from the Home Office. Once the drink drive accidents have been identified using Coroners and STATS9 data, then the resulting casualties in these accidents are identified from STATS9 data. Completeness of data and reliability of estimates Both sources of data from the Police and Coroners on drink drive accidents are incomplete. In recognition of the uncertainty associated with the estimates produced from this data the numbers of accidents and casualties are rounded to the nearest throughout this article. 39

44 In the case of the STATS9 breath test data, some drivers and riders are not breath tested due to it not being possible to administer a test. Some drivers and riders not tested might have failed if a test could have been administered. Probably as a result of ACPO s policy, the percentage of drivers tested increased dramatically between 995 and 999, whereas prior to 996 less than a third of drivers involved in injury accidents were tested. By 998 this proportion had risen to over half and remains at that level. For many drivers or riders killed in road accidents, a post-mortem blood alcohol level is not available, either because the casualty died more than twelve hours after the accident, no test was carried out, or because some of the data are not reported to the Department by Coroners and Procurators Fiscal. Adjustments to the reported data are therefore required to produce a more reliable estimate of the actual number of drink drive accidents and their related casualties. The estimates published here are based on a method described by Derek Jones in the 989 edition of Road Accidents Great Britain (RAGB). This method has two parts: a) the number of fatal accidents where a driver or rider died with an illegal alcohol level is estimated from the Coroners and Procurators Fiscal data. b) the number of accidents where a surviving driver or rider had an illegal alcohol level is estimated from data, based on a calculation of the proportion of these alcohol related accidents which can be identified from the STATS9 breath test data. Part b) was revised in 993 in the light of research by Dr J Broughton of the Transport Research Laboratory (TRL), published in TRL Report PR4 The Actual Number of Non-Fatal Drink drive Accidents. This provided a method which takes into account the fact that relatively more of the drivers and riders involved in fatal and serious accidents are breath tested than in slight accidents, whereas previously a single factor had been used to allow for under-reporting for all accident severities. The revised estimates were first published in RAGB 992. Estimates for 28 are provisional. As coroners data are available for analysis a year later than the main road accident data, final estimates can only be made eighteen months in arrears. Around 57 per cent expected to be available ultimately were available for inclusion in the provisional estimates in this article. The provisional estimates for serious and slight accidents depend on breath test data and do not change in the final estimates. The Coroners data affect only the numbers of casualties from fatal accidents and these form a small proportion of serious and slight casualties. The estimates for fatalities depend mainly on coroners data and are particularly susceptible to revision between the provisional and final figures. 4

45 4. Contributory factors to reported road accidents Christopher Waite, Road Safety Research and Statistics, Department for Transport Summary This article describes the scope and limitations of the information on contributory factors collected as part of the national road accident reporting system, and presents results from the fourth year of collection with a focus on the two factors related to speed. Failed to look properly was again the most frequently reported contributory factor and was reported in 37 per cent of all accidents reported to the police in 28. Four of the five most frequently reported contributory factors involved driver or rider error or reaction. For fatal accidents the most frequently reported contributory factor was loss of control, which was involved in 32 per cent of fatal accidents. Fourteen per cent of accidents had a speed related contributory factor reported, either exceeding the speed limit or travelling too fast for conditions. This rose to 24 per cent for fatal accidents, accounting for 25 per cent of all road deaths. Twenty three per cent of fatalities in these accidents were motorcyclists. Young drivers were more likely to have a speed related contributory factor reported than those over 25, and more than four times as many male drivers had a speed factor reported as female drivers. Forty one per cent of male fatalities aged 6 25 were in accidents where a speed factor was reported. Introduction From 25 all police forces in Great Britain have been reporting contributory factors as an integral part of the STATS9 collection system. The contributory factors system has been developed to provide some insight into why and how road accidents occur. Contributory factors are designed to give the key actions and failures that led directly to the actual impact to aid investigation of how accidents might be prevented. The factors are largely subjective, reflecting the opinion of the reporting police officer, and are not necessarily the result of extensive investigation. Some factors are less likely to be recorded since evidence may not be available after the event. While this information is valuable in helping to identify ways of improving safety, care should be taken in its interpretation. Part of this article presents general analysis from accidents reported to the police in 28 and explains the scope of the system, along with the limitations of its use. However, much of the value of this data is in assessing what happens in particular types of accident or comparing factors for different groups. Part 2 looks at the factors exceeding speed limit and travelling too fast for conditions and the characteristics of accidents involving these two factors. 4

46 Part Contributory factor system and general analysis Contributory factor data The contributory factor system allows the recording of up to six factors in those accidents reported at the scene by the police. Multiple factors may be recorded against an individual participant in the accident, either a vehicle, a casualty or an uninjured pedestrian. Factors relating to a driver/rider should be assigned to their vehicle. Any given factor may be assigned to a number of participants. Both accidents and vehicles can have more than one contributory factor attributed to them, therefore percentages in this article will not necessarily add up to. On average 2.4 contributory factors per accident were reported in 28. The form used by the police to report contributory factors can be found towards the rear of this publication (see contents page). The form includes the full list of all 77 contributory factors used by the police. The contributory factors are largely subjective and depend on the skill and experience of the investigating officer to reconstruct the events which led directly to the accident. They reflect the reporting officer s opinion at the time of reporting and are not necessarily the result of extensive investigation. Furthermore, it is recognised that subsequent enquiries could lead to the reporting officer changing his opinion. The contributory factors are therefore different in nature from the remainder of the STATS9 data which is based on the reporting of factual information. This should be kept in mind when interpreting the data. It is important to note that where some factors may have contributed to the cause of an accident it may be difficult for a police officer attending the scene after the accident has occurred to identify these factors. As a result some contributory factors may be less likely to be reported. For instance, while factors such as emergency vehicle on a call or defective traffic signals may be more obvious for a police officer attending the scene and so may be reported with some confidence, for other factors, such as exceeding speed limit or driver nervous, uncertain or panicked, it may not always be possible for the police officer to identify whether these factors took place and contributed to the accident. In addition, contributory factors are disclosable in court and police officers would require some supporting evidence before reporting certain factors. It is also important to note that not all reported accidents are included in the following analysis of the contributory factors data. For accidents in which a police officer did not attend the scene it may not be possible for the reporting officer to accurately report the correct contributory factors. Accidents in which no contributory factors were reported are also excluded. Seventy seven per cent of accidents reported to the police in 28 meet these criteria to be included in the contributory factor analysis. This includes 93 per cent of fatal accidents and 88 per cent of serious accidents. There is some variation in what proportion of different vehicle types are included, with more than five out of every six heavy goods vehicles and motorcycles involved in accidents in 28 being included in this analysis, compared with less than two thirds of pedal cycles and buses or coaches. These figures are consistent with those from 26 and 27. More information about accidents included in the contributory factor analysis can be found in the web tables in the Reported Road Casualties Great Britain: 28 section at gov.uk/pgr/statistics/datatablespublications/accidents/casualtiesgbar/. 42

47 28 results Each of the 77 contributory factors fits into one of nine categories. Chart 4a shows the percentage of accidents reported to the police with contributory factors in each category. The contributory factor category driver/rider error or reaction was the most frequently reported category, involved in 68 per cent of all accidents reported to the police, as in 27. It was also the most frequently reported type for each severity of accident. Injudicious action (including travelling too fast for conditions, following too close and exceeding speed limit) was the second most frequently reported category, involved in 25 per cent of all accidents. However this increases to 29 per cent of fatal accidents. Special codes (including stolen vehicle, vehicle in course of crime and emergency vehicle on a call) were reported for 4 per cent of all accidents. Pedestrian contributory factors, which are those where the factor has been attributed to an injured or uninjured pedestrian involved in the accident, were reported in 3 per cent of all accidents and 9 per cent of fatal accidents. Chart 4a: Contributory factor type: Reported accidents by severity: GB 28 Table 4a shows the percentage of accidents in which each contributory factor was reported, including a breakdown by accident severity. Failed to look properly was the most frequently reported contributory factor and was involved in 37 per cent of all reported accidents. This was followed by failed to judge other person s path/speed (9 per cent) and careless, reckless or in a hurry (5 per cent). Failed to look properly was the most frequently reported contributory factor for slight and serious accidents (38 per cent and 3 per cent). For fatal accidents the most frequently reported contributory factor was loss of control, which was involved in 32 per cent of fatal accidents. Loss of control was also the second largest contributory factor for serious accidents (2 per cent). Four of the five most frequently reported contributory factors were some kind of driver/ rider error or reaction, which includes failed to look properly and failed to judge other person s path or speed. 43

48 Table 4a: Contributory factors: Reported accidents by severity: GB 28 Fatal accidents Serious accidents Slight accidents All accidents Per Per Per Per Contributory factor reported in accident Number cent 2 Number cent 2 Number cent 2 Number cent 2 Road environment contributed ,82 4 7, ,564 6 Poor or defective road surface Deposit on road (eg. oil, mud, chippings) , ,28 2 Slippery road (due to weather) 5 5, ,24 4,66 Inadequate or masked signs or road markings Defective traffic signals Traffic calming (eg. speed cushions, road humps, chicanes) Temporary road layout (eg. contraflow) Road layout (eg. bend, hill, narrow ,94 3 3,5 3 carriageway) Animal or object in carriageway 7 27,253,487 Vehicle defects ,69 2 2,574 2 Tyres illegal, defective or under inflated Defective lights or indicators Defective brakes Defective steering or suspension Defective or missing mirrors Overloaded or poorly loaded vehicle or trailer Injudicious action , , , Disobeyed automatic traffic signal ,35 2 2,338 2 Disobeyed Give Way or Stop sign or , ,35 3 markings Disobeyed double white lines Disobeyed pedestrian crossing facility Illegal turn or direction of travel Exceeding speed limit 33 4,52 7 5,27 5 7,3 5 Travelling too fast for conditions ,9 9, ,28 9 Following too close ,62 7 8,95 6 Vehicle travelling along pavement Cyclist entering road from pavement ,28 Driver/rider error or reaction, , , , Junction overshoot ,64 2 3,76 2 Junction restart (moving off at junction) 7 245, ,4 2 Poor turn or manoeuvre 23 2,77 3 5,66 4 8,3 4 Failed to signal or misleading signal ,24 2 2,493 2 Failed to look properly , , ,34 37 Failed to judge other person s path or speed , ,3 2 25,343 9 Passing too close to cyclist, horse rider or ,569,93 pedestrian Sudden braking , ,292 7 Swerved , ,963 4 Loss of control ,55 2 4,87 4 9,576 5 Impairment or distraction , ,59 5,562 2 Impaired by alcohol 237, ,36 5 6,758 5 Impaired by drugs (illicit or medicinal) Fatigue ,374,82 Uncorrected, defective eyesight Illness or disability, mental or physical ,356,848 Not displaying lights at night or in poor visibility Cyclist wearing dark clothing at night Driver using mobile phone Distraction in vehicle ,46 2 2,84 2 Distraction outside vehicle ,65 2,93 ( continued) 44

49 Table 4a: (continued) Fatal accidents Serious accidents Slight accidents All accidents Per Per Per Per Contributory factor reported in accident Number cent 2 Number cent 2 Number cent 2 Number cent 2 Behaviour or inexperience , , ,58 24 Aggressive driving 82 8,57 5 3, , 4 Careless, reckless or in a hurry ,95 5 6, ,237 5 Nervous, uncertain or panic , ,8 2 Driving too slow for conditions or slow vehicle (eg tractor) Learner or inexperienced driver/rider 5, 5 6,6 6 7,282 6 Inexperience of driving on the left Unfamiliar with model of vehicle ,6 Vision affected by: 57 7,928 9,64 3,726 Stationary or parked vehicle(s) , ,239 3 Vegetation Road layout (eg. bend, winding road, hill crest) 22 28,563,866 Buildings, road signs, street furniture Dazzling headlights Dazzling sun ,55 2 2,935 2 Rain, sleet, snow, or fog ,26 2 2,62 2 Spray from other vehicles Visor or windscreen dirty or scratched Vehicle blind spot 25 2,328,564 Pedestrian only (casualty or uninjured) , ,425 7,75 3 Pedestrian crossing road masked by , ,75 3 stationary or parked vehicle Pedestrian failed to look properly ,22 5 9, ,75 Pedestrian failed to judge vehicle s path or ,73 3 3,845 3 speed Pedestrian wrong use of pedestrian crossing ,37 facility Dangerous action in carriageway (eg. playing) ,3,44 Pedestrian impaired by alcohol , ,494 2 Pedestrian impaired by drugs (illicit or medicinal) Pedestrian careless, reckless or in a hurry 53 2,9 6 3,79 3 5,35 4 Pedestrian wearing dark clothing at night Pedestrian disability or illness, mental or physical Special codes 33 6,5 5 4,54 4 5,678 4 Stolen vehicle Vehicle in course of crime Emergency vehicle on a call Vehicle door opened or closed negligently Other ,65 2 3,35 3 Total number of accidents 2,7 2,424 8,988 3,582 Includes only accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Columns may not add up to per cent as accidents can have more than contributory factor. Table 4b compares the ten most frequently reported contributory factors in 25, 26, 27 and 28. The ten factors remained the same in all three years; however there were some small changes in the order and frequency of the factors. The largest change was an increase in the most frequently reported factor, failed to look properly, which was reported in 32 per cent of accidents in 25 and 37 per cent in 28. At this stage it is not possible to tell whether changes are the result of the reporting police officers developing their understanding of the new system or a genuine change in the kinds of factors that contribute to accidents. 45

50 Table 4b: Reported contributory factors: GB 25/26/27/28 comparison Per Per Per Per Contributory factor reported in accident 2 Number cent Number cent Number cent Number cent Failed to look properly 46, , , ,34 37 Failed to judge other person s path or speed 26, , , ,343 9 Careless, reckless or in a hurry 23, , , ,237 5 Loss of control 2,24 4 2, ,54 5 9,576 5 Poor turn or manoeuvre 22,52 5 2,6 4 9, ,3 4 Slippery road (due to weather) 4,268 3, ,54 4,66 Pedestrian failed to look properly 3,69 9 3,879 3, ,75 Travelling too fast for conditions 7,7 2 6,8 3,856 2,28 9 Sudden braking,273 7, ,99 7 9,292 7 Following too close,847 7,24 7 8, ,95 6 Total number of accidents 47,59 45,798 4,36 3,582 Includes only accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Includes only the ten most frequently reported contributory factors. Table 4c shows, for each vehicle type, the percentage of vehicles which had each contributory factor. The table shows the ten most frequently reported contributory factors for each vehicle type. The percentages in this table are different from those in Table 4a which gives the percentage of accidents with each contributory factor. For example when looking at Failed to look properly 5,33 vehicles had this contributory factor out of a total of 24,55 vehicles (2 per cent of vehicles). The vehicles which had this contributory factor were in 48,34 accidents out of a total of 3,582 reported accidents (37 per cent of accidents). Part of the reason for the lower number when looking at the percentage of vehicles is that,79 vehicles (42 per cent) involved in accidents had no contributory factor reported. Failed to look properly was the most frequently reported contributory factor for every vehicle type except motorbikes. This factor was analysed in some detail in the 26 publication. Failed to judge other person s path or speed was the second most frequently reported factor for cars and goods vehicles and was reported for per cent of vehicles overall. Motorcycles had a notably higher percentage of the contributory factors loss of control (5 per cent) and learner/inexperienced driver (9 per cent) when compared to other vehicle types. Sudden braking was the second most frequently reported contributory factor for buses or coaches (3 per cent). Cyclist entering road from pavement was attributed to 9 per cent of pedal cycles in accidents and cyclist wearing dark clothes at night was attributed to 4 per cent. Seven per cent of heavy goods vehicles (HGVs) involved in accidents had vehicle blind spot as a contributory factor. Exceeding speed limit was attributed to 3 per cent of cars involved in accidents, while travelling too fast for conditions was attributed to 5 per cent. For fatal accidents these figures are both 8 per cent. 46

51 Table 4c: Contributory factors: Vehicles in reported accidents by vehicle type: GB 28 Pedal cycles Motorcycles Cars Bus or Coach Contributory factor attributed to vehicle 3 Number Per cent Number Per cent Number Per cent Number Per cent Failed to look properly Failed to judge other person s path or speed Careless, reckless, in a hurry Loss of control Poor turn or manoeuvre Slippery road (due to weather) Travelling too fast for conditions Sudden braking Following too close Learner or inexperienced driver/rider Exceeding speed limit Impaired by alcohol Vision affected by stationary or parked vehicle Junction overshoot Passing too close to cyclist, horse rider, pedestrian Vehicle blind spot Cyclist entering road from pavement Cyclist wearing dark clothing at night No lights at night or in poor visibility 2, , ,83 2,,678 2,888,677,527,24,45 725, , 2,54 6,74 5, 4,4 2,49,9 7,32 6,79 5,293 5,878 5,646 3,799 2,447, Vehicles with no contributory factor 5, , , , Number of vehicles,782 8,88 85,326 5,5 Light goods vehicle Heavy goods vehicle All vehicles 2 Contributory factor attributed to vehicle 3 Number Per cent Number Per cent Number Per cent Failed to look properly Failed to judge other person s path or speed Careless, reckless, in a hurry Loss of control Poor turn or manoeuvre Slippery road (due to weather) Travelling too fast for conditions Sudden braking Following too close Learner or inexperienced driver/rider Exceeding speed limit Impaired by alcohol Vision affected by stationary or parked vehicle Junction overshoot Passing too close to cyclist, horse rider, pedestrian Vehicle blind spot Cyclist entering road from pavement Cyclist wearing dark clothing at night No lights at night or in poor visibility 2,58,44, ,76, ,33 26,777 2,583 9,66 8,535 4,777 2,534 9,874 8,928 7,36 7,28 6,62 4,736 3,89,93,577, Vehicles with no contributory factor 4, ,47 4,79 42 Number of vehicles,574 7,734 24,55 Includes only vehicles in road accidents where a police officer attended the scene and in which a contributory factor was reported. Columns may not add up to per cent as accidents can have more than one contributory factor. 2 Includes other vehicles types and cases where the vehicle type was not reported. 3 Includes only the ten most frequently reported contributory factors for each vehicle. 47

52 Table 4d shows the most frequent pairs of contributory factors assigned to the same vehicle or pedestrian casualty in road accidents reported to the police in 28. The pair of contributory factors most frequently reported for the same vehicle were failed to look properly and failed to judge other person s path or speed, with 6 per cent of vehicles having both factors assigned to them. This means that over half of all vehicles that were assigned failed to judge other person s path or speed were also assigned failed to look properly. These were also the two most frequently reported contributory factors in all accidents. The pair of contributory factors most frequently assigned to the same pedestrian casualty were pedestrian failed to look properly and pedestrian careless, reckless or in a hurry. Eighteen per cent of pedestrian casualties were assigned this pair of factors. Over 8 per cent of all pedestrian casualties with pedestrian careless, reckless or in a hurry as a contributory factor were also assigned pedestrian failed to look properly. Table 4d: Most common pairs of contributory factors reported together,2 : GB 28 Factor with lower code 3 Factor with higher code 3 Number Per cent Vehicles Failed to look properly Failed to judge other person s path or speed 3,532 6 Poor turn or manoeuvre Failed to look properly 9,53 4 Failed to look properly Careless, reckless or in a hurry 8,726 4 Poor turn or manoeuvre Failed to judge other person s path or speed 4,45 2 Slippery road (due to weather) Loss of control 4,243 2 Travelling too fast for conditions Loss of control 3,967 2 Poor turn or manoeuvre Careless, reckless or in a hurry 3,669 2 Failed to judge other person s path or speed Careless, reckless or in a hurry 3,6 Slippery road (due to weather) Travelling too fast for conditions 3,56 Loss of control Careless, reckless or in a hurry 3,4 Following too close Failed to look properly 2,87 Following too close Failed to judge other person s path or speed 2,696 Disobeyed Give Way or Stop sign or markings Failed to look properly 2,632 Travelling too fast for conditions Careless, reckless or in a hurry 2,45 Swerved Loss of control 2,375 Exceeding speed limit Loss of control 2,28 Loss of control Learner or inexperienced driver/rider,998 Travelling too fast for conditions Failed to look properly,986 Exceeding speed limit Careless, reckless or in a hurry,924 Aggressive driving Careless, reckless or in a hurry,98 All vehicles in accidents 24,55 Pedestrian casualties Pedestrian failed to look properly Pedestrian careless, reckless or in a hurry 3,827 8 Pedestrian crossing road masked by stationary or parked Pedestrian failed to look properly 2,799 3 vehicle Pedestrian failed to look properly Pedestrian failed to judge vehicle s path or speed 2,7 3 Pedestrian failed to look properly Pedestrian impaired by alcohol,244 6 Pedestrian failed to judge vehicle s path or speed Pedestrian careless, reckless or in a hurry,22 6 All pedestrian casualties in accidents 2,33 Includes only participants in accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Includes the 2 pairings most frequently reported to vehicles and the 5 most frequently reported to pedestrian casualties. 3 All contributory factors are recorded by a code number between and 999. The factor with the lower code number is listed first. 48

53 Part 2 Accidents involving speed as a contributory factor This part of the article gives some information on accidents involving two contributory factors which relate to speed - exceeding the speed limit and travelling too fast for the conditions. As stated in part of this article, the contributory factors are the opinions of police officers based on the evidence available to them at the time of reporting the accident. As a result some contributory factors may be less likely to be reported. These two factors relating to speed may be difficult to determine after the event, especially in less serious accidents. Other factors, which are easier to determine after the event, may also indicate the involvement of inappropriate or excessive speed. These include loss of control, following too close and sudden braking. The definitions of the two factors considered are: Exceeding the speed limit should be reported when the driver caused, or contributed to the accident by exceeding the posted speed limit. It should also be reported when the actions of another road user were the immediate cause of the accident but a speeding vehicle also contributed to the collision. Travelling too fast for conditions should be reported when the driver was travelling within the speed limit, but their speed was not appropriate for the road conditions and/or vehicle type, and contributed to the accident. For the purpose of this part of the article accidents in which both exceeding the speed limit and travelling too fast for the conditions were reported will be counted under exceeding the speed limit only. It should be noted that if a driver was exceeding the speed limit and travelling too fast for the conditions, reporting officers are asked to report only the former factor. However in a number of cases both factors are reported; these drivers will only be counted as exceeding the speed limit. As a result the figures for travelling too fast for conditions in this section are different from the figures in Part. Over the four years that contributory factor data has been recorded there has been a fall in the proportion of accidents in which travelling too fast for conditions is reported, particularly in fatal accidents, while exceeding the speed limit has seen a slight rise in reporting. At this stage it is not possible to tell whether changes are the result of the reporting police officers developing their understanding of the new system or a genuine change in these factors contributing to accidents. 49

54 Table 4e shows the number and percentage of accidents in which the two contributory factors were reported in 28. Exceeding the speed limit was reported as a contributory factor in 5 per cent of all accidents. However, the factor became more significant with the severity of the accident. It was reported in 4 per cent of fatal accidents and these accidents accounted for 362 fatalities, 5 per cent of all deaths. Accidents that involved exceeding the speed limit as a contributory factor were three times as likely to result in at least one fatality compared to those that did not. The factor travelling too fast for conditions was a contributory factor in 8 per cent of accidents. Again the proportion of accidents where it was reported rose with the severity of the accident and 9 per cent of fatal accidents involved it as a factor, resulting in 224 fatalities, per cent of all deaths. Table 4e: Speed as a contributory factor: Reported accidents and casualties by severity : GB 28 Accidents Fatal Serious Slight Total Contributory factor in accident Number Per cent Number Per cent Number Per cent Number Per cent Exceeding speed limit 33 4,52 7 5,27 5 7,3 5 Travelling too fast for conditions , ,77 8,682 8 Exceeding speed limit or travelling , , ,785 4 too fast for conditions Total number of accidents 2,7 2,424 8,988 3,582 Casualties Killed Seriously injured Slightly injured Total Contributory factor in accident Number Per cent Number Per cent Number Per cent Number Per cent Exceeding speed limit 362 5, ,95 6,392 6 Travelling too fast for conditions ,64 9 4,2 9 6,58 9 Exceeding speed limit or travelling , , ,9 5 too fast for conditions Total number of casualties 2,35 23,95 58,669 84,25 Includes accidents and casualties in accidents where a police officer attended the scene and a contributory factor was reported. 2 Excluding accidents and casualties in accidents which had exceeding the speed limit reported as a contributory factor. 5

55 In 28 there were just under 28, casualties in accidents where a speed contributory factor was reported. These included 586 fatalities and over 4, serious injuries, accounting for 25 per cent of road deaths and 8 per cent of serious injuries. Table 4f shows how many casualties were involved in accidents where a speed factor was reported by road user group and severity. For all the main road user groups the proportion of casualties that were killed or seriously injured was higher in accidents where one of the speed factors was reported. In accidents where a vehicle was exceeding the speed limit one in ten casualties was a motorcyclist, but they made up over a quarter of fatalities. Similarly pedestrians made up one in nine fatalities despite making up one in every twenty seven casualties. In accidents where a vehicle was travelling too fast for conditions there was a similar pattern. Motorcyclists made up 8 per cent of all casualties and pedestrians 2 per cent, but of those casualties that were killed 8 per cent were motorcyclists and 7 per cent pedestrians. In accidents with a speed contributory factor reported 44 per cent of motorcyclist casualties were killed or seriously injured, compared to 28 per cent in other accidents. Twenty nine per cent of motorcyclist deaths were in accidents where a speed factor was reported. Only 4 per cent of pedestrian casualties were in accidents where a speed factor was reported. However, 42 per cent of pedestrians in accidents with a speed factor were killed or seriously injured, compared to 27 per cent of those in other accidents. Table 4f: Casualties by road user group and severity in accidents with a speed factor reported :GB 28 Pedestrians Pedal cyclists Motorcyclists Car occupants All road users 2 Casualties in accidents with a speed ,297 22,92 27,9 factor 3 Of which KSI ,5 3,73 4,685 Of which killed Percentage KSI Percentage killed Casualties in all accidents 2,33,479 8,323 23,73 84,25 Of which KSI 4 5,75,952 5,59,95 25,546 Of which killed ,66 2,35 Percentage KSI Percentage killed 3 2 Includes casualties in accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Includes other road user types. 3 Casualties in accidents with either exceeding speed limit or travelling too fast for conditions reported as a contributory factor. 4 Killed or seriously injured. 5

56 Speed contributory factors by age and gender Speed factors associated with drivers vary depending on their age and gender. Chart 4b shows the percentage of drivers involved in accidents reported to the police in 28 that were reported as exceeding the speed limit or travelling too fast for conditions by age and gender. The number of male drivers reported as exceeding the speed limit was more than six times the number of female drivers. More than three times as many male drivers were assigned the factor travelling too fast for conditions than female drivers. Overall more than four times as many male drivers had a speed factor reported than female drivers. Younger drivers were more likely to be assigned speed factors, with 6 per cent of 6 25 year olds involved in accidents exceeding the speed limit and 8 per cent travelling too fast for conditions. Both of these rates fell as the age of the driver increased. Chart 4b: Proportion of reported drivers allocated speed factors by driver age and gender: GB 28 52

57 Table 4g gives a breakdown of casualties in accidents reported to the police with a speed contributory factor recorded by age and gender, and shows what percentage of all casualties of that age and gender were involved in these accidents. In 28, men made up 63 per cent of casualties in accidents where a speed factor was reported, while 37 per cent of casualties were women. However, 74 per cent of serious injuries and 8 per cent of fatalities in these accidents were male. Accidents with a speed factor accounted for 6 per cent of all male road casualties and 3 per cent of all females. This percentage rose with severity and 27 per cent of male fatalities and 9 per cent of female fatalities were in accidents with a speed factor. Casualties aged 6 to 25 were more likely to be involved in an accident with a speed factor than other age groups. Twenty one per cent of casualties in this age group were in accidents with a speed factor, compared to just 8 per cent of casualties over 65. This trend was followed in all severities, and 4 per cent of male road accident fatalities aged 6 25 were in accidents where a speed factor was reported. Child casualties (aged 5) made up 6 per cent of casualties in accidents with a speed factor, compared to 9 per cent of casualties in all accidents. There were 2 children killed in accidents where a speed factor was reported, 8 per cent of all child fatalities. Table 4g: Casualties in reported accidents with a speed contributory factor reported by severity, age and gender,2 : GB 28 Killed Killed or seriously injured All casualties Age group Male Female Total 3 Male Female Total 3 Male Female Total 3 Number of casualties , , ,98 7,64 3,774, ,24 4,84 2,546 7, ,647 2,385 6, ,9 All ages ,486,97 4,685 7,73,76 27,9 Percentage of all casualties of that severity, age and gender in accidents with a speed factor All ages Includes casualties in accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Casualties in accidents with either exceeding speed limit or travelling too fast for conditions reported as a contributory factor. 3 Includes cases where gender was not recorded. 4 Includes cases where age was not recorded. 53

58 Speed contributory factors by road type Table 4h shows how many casualties were in accidents reported to the police where a speed contributory factor was reported on different types of road in 28. Accidents on rural roads were more likely to have a speed contributory factor reported than those on motorways or urban roads, with 7 per cent of casualties on rural A roads and 23 per cent on minor rural roads occurring in accidents where one of the factors was reported, compared to 2 per cent on urban roads and 4 per cent on motorways. In particular 26 per cent of casualties on minor rural roads with a 6 mph speed limit were in accidents in which a speed factor was reported. Only 3 per cent of casualties in accidents where a vehicle was exceeding the speed limit happened on motorways, while 58 per cent were on urban roads and 39 per cent on rural roads. Motorways had a lower proportion of accidents with this factor than other road classes. Of accidents where a vehicle was travelling too fast for conditions, 7 per cent happened on motorways, 36 per cent on urban roads and 57 per cent on rural roads. Motorways had a higher proportion of accidents with this factor than other road classes. Table 4h: Casualties in reported accidents with a speed contributory factor by road type,2 : GB 28 Urban roads Rural roads Speed Limit Motorways A roads Minor roads 3 A roads Minor roads 3 All roads 4 Number of casualties 2 mph or less mph 23 4,42 6,48 768,749 3,63 4 mph ,633 5 mph ,8 6 mph ,52 4,649 8,44 7 mph,28 64, ,52 All limits,48 5,656 6,977 6,597 7,25 27,9 Percentage of all casualties on that road type that were in accidents with a speed factor 2 mph or less mph mph mph mph mph All limits Includes casualties in accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Accidents with either exceeding speed limit or travelling too fast for conditions reported as a contributory factor. 3 B, C and unclassified roads. 4 Including those not recorded as urban or rural. 54

59 Speed contributory factors by time of day, week and year The proportion of accidents that were assigned a speed factor varied between months. In 28 exceeding the speed limit was recorded in a higher proportion of accidents in spring and summer than it was in autumn and winter, ranging between 4 per cent of accidents in January and 6 per cent of accidents in May. Conversely, travelling too fast for conditions was reported in a higher proportion of accidents in the winter. It was reported in per cent of accidents in November, compared to 7 per cent of accidents in June. In 28 3 per cent of accidents that were assigned a speed factor occurred on a Saturday or Sunday, and one of the two factors was reported in 7 per cent of accidents at weekends, compared to 2 per cent of accidents on weekdays. Chart 4c shows the proportion of accidents that were assigned a speed factor in 28 by time of day. Speed factors were reported in a higher proportion of accidents at night than during the day. Between pm and 6am at least one of the factors was reported in 23 per cent of accidents, compared to per cent between 8am and 6pm. Exceeding the speed limit was reported in 2 per cent of accidents between pm and 5am, the only time of day when it was reported more often than travelling too fast for conditions. Between 8am and 6pm it was reported in just 4 per cent of accidents. There was a smaller difference between the amount travelling too fast for conditions was reported at night and during the day, with per cent of accidents between pm and 6am having it reported compared to 7 per cent of accidents between am and 5pm. Unlike exceeding the speed limit, travelling too fast for conditions peaked at either end of the night rather than in the middle, with per cent of accidents between pm and midnight and 2 per cent of accidents between 4am and 7am having the factor reported. Chart 4c: Proportion of accidents with speed factors reported by time of day: GB 28 55

60 Vehicles with speed contributory factors assigned Table 4i shows how many times the two speed factors were allocated to different vehicle types. Motorcycles were the vehicle type most likely to have one of the speed factors assigned to them in 28. Five per cent were reported as exceeding the speed limit compared to 3 per cent of cars and 6 per cent of motorcycles were travelling too fast for conditions compared to 5 per cent of cars. Heavy goods vehicles were less likely to have either of the speed factors assigned than other vehicles. Only 4 per cent of heavy goods vehicles had one of the factors assigned to them. Light goods vehicles were less likely to be exceeding the speed limit than cars, with 2 per cent of light goods vehicles involved in accidents having the factor reported. A higher proportion of vehicles were assigned travelling too fast for conditions than exceeding the speed limit for all vehicle types. Table 4i: Vehicles with speed factors reported by vehicle type : GB 28 Motorcycles Cars Light goods vehicles Heavy goods vehicles All vehicles 2 Contributory factor attributed to vehicle Number Per cent Number Per cent Number Per cent Number Per cent Number Per cent Exceeding speed limit , ,28 3 Travelling too fast for conditions 3,9 6 8, ,944 5 Exceeding speed limit or travelling too fast for conditions 2,3 4, ,52 8 All vehicles in accidents 8,88 85,326,574 7,734 24,55 Includes vehicles in accidents where a police officer attended the scene and in which a contributory factor was reported. 2 Includes other vehicle types. 3 Excluding vehicles which had exceeding the speed limit reported as a contributory factor. 56

61 In 28 almost seven out of eight vehicles with one of the speed contributory factors reported had at least one other contributory factor reported. Loss of control was assigned to 3 per cent of vehicles with a speed factor and careless, reckless or in a hurry to 2 per cent. Chart 4d shows the association between the two speed factors and other contributory factors. Exceeding the speed limit tended to have the most association with factors related to other illegal activities. Twenty four per cent of vehicles assigned vehicle in course of crime also had the factor exceeding the speed limit, as did 2 per cent with stolen vehicle and 6 per cent with impaired by alcohol. The contributory factors that travelling too fast for conditions had the highest association with tended to be those related to the road environment and to bad weather. For example it was reported for 22 per cent of vehicles that were assigned slippery road (due to weather). Chart 4d: Percentage of vehicles with selected contributory factor that had a speed factor reported: GB 28 57

62 5. Comparing police data on road accidents with other sources Matthew Tranter, Road Safety Research and Statistics, Department for Transport Summary Comparisons with death registrations show that very few, if any, road accident fatalities are not reported to the police. It has long been known that a considerable proportion of non-fatal casualties are not known to the police and hospital, survey and compensation claims data all indicate a higher number of casualties than are reported. Our best current estimate, derived from survey data with cross-checking against other data sources, is that the total number of road casualties in Great Britain each year, including those not reported to police, is within the range 68 thousand to 92 thousand with a central estimate of 8 thousand. Part 2 of this article contains further details of these estimates. Police data on road accidents (STATS9), whilst not perfect, remains the most detailed, complete and reliable single source of information on road casualties covering the whole of Great Britain, in particular for monitoring trends over time. However, both hospital and survey data are likely to provide further useful evidence on trends in the future. Introduction For many years the police have provided data on road accidents reported to them involving casualties under the STATS9 system. This source provides almost all the data in this publication. What other sources of data are there on road accidents and casualties? Besides the police data, there are a number of further sources of data relating to road accident casualties, including hospital, survey and compensation claims databases. Part of this article provides an overview of a Death registrations data number of these sources, focusing on Hospital Episode Statistics: Government datasets with national coverage, inpatient admissions and broadly arranged by the severity of the Hospital Episode Statistics: A&E casualties included (see box). attendances DWP Compensation Claims data In addition to these datasets, there are a National Travel Survey data on number of other sources of data on road road accidents casualties, for example in-depth accident investigation studies such as the On The Spot study outlined in article 7. These sources typically include much more detail than is contained in any of those covered in this article, but cover only a small, often nonrepresentative sample of accidents and casualties and for this reason will not be included here. 58

63 How do these data sources add to the overall picture? Although STATS9 is the most detailed and useful source of information on road casualties at national level, it is not a complete or perfect dataset. It is therefore desirable to use complementary sources to build a balanced and comprehensive picture of the nature and extent of road accidents. Other datasets can be useful both as a check on the quality and completeness of STATS9 and in providing information which is not collected by the police, for example relating to more detailed medical consequences of road accidents. Making comparisons of STATS9 with other data sources is not straightforward, as there are often differences in definitions and changes in data collection and recording practices which can affect trends over time. In addition, some of the data collections are relatively new and a longer time series is needed before a full analysis can be carried out. Details of what each source adds, and strengths and limitations, are given in part of this article. What is the best estimate of the total number of road casualties in Great Britain? From 27, the Department s National Travel Survey (NTS) has asked about involvement and injury in road accidents. Although this is based on a sample of the population, and therefore subject to sampling variability, it is the only source providing complete coverage of casualties (particularly those who do not report an accident to police or attend hospital). Grossing up the survey estimate to the population suggests that the total number of road casualties in Great Britain is between 68 thousand and 92 thousand per year, with a best estimate of around 8 thousand. This is more than three times the number of casualties that are recorded in STATS9. Of this total, we estimate that very roughly 8 thousand people are seriously injured each year, using the same broad definition of a serious injury that is used in STATS9. The NTS also gives an estimate of 35 thousand people who attend hospital following an accident. A discussion of how these estimates have been derived, and their limitations, is given in part 2 of this article. Does this mean that the police data are not reliable or useful? No. STATS9 remains the single most useful source of data on road accidents and resulting casualties in Great Britain. In particular, it is the only national source to provide detailed information on accident circumstances, vehicles involved and resulting casualties. However, as has long been known, STATS9 is not a complete record of all injury accidents and resulting casualties, and this should be borne in mind when using and analysing the data. The above estimates illustrate this. 59

64 What about trends over time? Although STATS9 does not provide complete coverage of road accidents and casualties, this does not in itself make it unsuitable for monitoring changes over time, assuming that levels of reporting to police have not changed. There have been a number of studies of levels of reporting of road accidents in recent years (see part 2 of this article for references). These have shown different patterns and to date there is no clear or conclusive evidence of a systematic change in levels of reporting at national level. In addition: Most, if not all, road accident fatalities are included in the police data, which shows a fall of 29 per cent between the average and 28. This reduction in the number of people dying in road accidents is confirmed by death registrations data. Both police and hospital admissions data show reductions in more severely injured road casualties (though the size of the reduction shown by the police data is greater). It should be noted that the Health and Social Care Information Centre publish a warning regarding the use of hospital data for trend analysis. For what purposes are the police data useful, and what do users need to bear in mind? As outlined above, STATS9, although incomplete, remains the single best source of information on accident circumstances and vehicles involved in personal injury accidents. Uses of the data include development and monitoring of road safety initiatives at both local and national level, developing and evaluating legislative changes, targeting road safety publicity campaigns, development of road and vehicle engineering measures and identifying public health issues related to road safety. Users of STATS9 data should be aware that, whilst comparisons with survey data (see part of this article) suggest that the police data is sufficiently representative of casualties by road user type, levels of reporting to police may vary according to the accident circumstances. For example, it is known that few single vehicle pedal cycle accidents are included in STATS9. In addition, particular local circumstances (for example organisational changes, reviews of coding practice and local initiatives) may affect the data and trends over time, particularly at local level. Therefore, as with most data sources, users of STATS9 are advised to carefully explore relevant issues before drawing conclusions from the data, and the Department is happy to offer advice in this area. See and then choose understanding the data and why are there fluctuations? 6

65 Part : Sources of road accident data STATS9 data Data source and coverage Accidents reported to and recorded by police forces in Great Britain. Covers personal injury accidents and resulting casualties occurring on the public highway and involving at least one vehicle. Casualties are coded as killed, seriously injured or slightly injured (see definitions) Strengths and limitations A well established source of data which has existed in a comparable format and with consistent injury definitions for several decades. Only includes accidents which become known to police, and therefore is an incomplete record of all road accident casualties as there is not a legal obligation to report all road accidents (see part 2) Trends shown by STATS9 will be influenced by any changes in propensity of incidents to be reported over time, and the proportion of reported accidents that are recorded by police. However, considerable efforts are made by DfT and the police to ensure continuity over time. Studies of levels of reporting to date have been inconclusive, but provide no evidence of a systematic change. Coding of injury severity made by police who are not medically trained and may not know full details at the time of an accident. Studies have shown police are more likely to underestimate injury severity, for example because of difficulties in identifying some types of injury at the accident scene. What it adds Historically the main source of numbers of and trends in road accidents and casualties the majority of the tables and analyses throughout this publication are derived from STATS9 data. Used to monitor progress towards casualty reduction targets Provides detailed information about accident circumstances and location, and vehicles involved, to inform development of road safety policies. Enables detailed analysis of accidents and casualties at local level Further information The form used by the police to record details of an accident can be found towards the end of this publication. Details of guidelines for completion of the STATS9 form can be found online: The notes section of this publication provides information on STATS9 coverage. STATS9 data is collected to an agreed national standard, and local and national government work closely with local police forces to achieve this. However, it has long been known that not all accidents become known to police, and are therefore not included in the figures presented in this publication. One reason for this is that there is no legal requirement to report an accident to police, provided that details are exchanged 2. 2 Legal requirement: 6

66 Death registrations data Data source and coverage Number of deaths derived from registrations of death certificates completed by a doctor or coroner, after processing by Office for National Statistics (ONS; for England and Wales) and General Register Office for Scotland (GROS). Cause of death coded using International Classification of Diseases th revision (ICD-) codes, allowing road transport accidents to be identified. Definitions are not identical to those used by police, in particular there is no time limit (in contrast to the 3 day cut-off used in STATS9 see definitions section) Based on date of death registration to 992 and from 26; based on date of death occurrence between 993 and 25. Strengths and limitations A well established and comprehensive data source with a long time series of data. Only covers fatalities a small proportion of total road accident casualties Known changes over time which affect trends (e.g. to coding of cause of death) No information on circumstances of the accident What it adds Provides strong evidence that most, if not all, road deaths become known to police Confirms that trends in fatalities recorded by police are reliable. Further information Table 5 of this publication provides numbers of registered deaths by age and sex ONS Mortality Statistics publication (covering England and Wales): Mortality statistics including the number of registered road deaths each year are published by the ONS and by GROS. Chart 5a shows registered road deaths compared with fatalities recorded in STATS9 both the level and trends are similar. Chart 5a: Comparison of STATS9 fatalities and registered road deaths: GB Source: STATS9 and Office for National Statistics/General Register Office Scotland 3 3 Data taken from previous editions of Road Casualties Great Britain; gaps relate to years when figures were not published due to unavailability of data. 62

67 Hospital Episode Statistics: Inpatient data Data source and coverage Hospital Episode Statistics (HES) record the number of inpatient admissions to hospitals in England, derived from hospital s patient administrative systems and collated by the Health and Social Care Information Centre. There are equivalent datasets for Scotland (Scottish Morbidity Record, SMR) and Wales (Patient Episode Database for Wales, PEDW). Records relate to episodes of care under a particular consultant. Excludes those attending Accident and Emergency (A&E) only but includes people admitted and discharged on the same day. Cause of injury coded using International Classification of Diseases th revision (ICD-) codes, allowing road traffic accidents to be identified. Strengths and limitations Provides information on medical consequences of accidents (not collected in STATS9), but does not have detailed information about accident circumstances. Coding of injury likely to be more accurate than in STATS9, but coding of location less accurate meaning some off-road incidents may be recorded as traffic accidents. Only covers casualties admitted to hospital, which will not include fatalities who die before admission, or those treated only in A&E, by GPs or elsewhere, or who receive no medical attention. However does include some casualties who do not report their accident to police. To date, not reliable for monitoring casualty trends over time due to changes in hospital practices and data collection. What it adds Indicates the public health burdens of road accidents around 4, hospital admissions a year in England. Information on nature of injuries sustained by road casualties. Matching to STATS9 offers scope to compare medical consequences with accident circumstances. Further information Article 6 in Road Casualties Great Britain 26 gives a more detailed comparison of HES and STATS9. See HES Online website: The Hospital Episode Statistics (HES) inpatient database, managed by the Health and Social Care Information Centre, contains data on inpatient admissions to hospitals in England. Each HES record contains clinical details of the patient s condition which allows the identification of patients whose injuries have been caused by a road traffic accident. 63

68 Comparing HES inpatient data with STATS9 The definition of seriously injured casualties in STATS9 4 includes anyone that is admitted to hospital, and also those with specific types of injury, whereas HES covers those admitted to hospital (regardless of any injury). The HES figures can be filtered to be broadly comparable with STATS9 5. Chart 5b shows comparisons between HES data and the STATS9 seriously injured category for England. Chart 5b: Comparison of STATS9 seriously injured casualties with hospital admissions: England 995/96 to 27/8 financial years Source: STATS9 and Hospital Episode Statistics, The NHS Information Centre for Health and Social Care Although the numbers of seriously injured casualties and the number of hospital admissions are similar, particularly in the mid-99s, many of those admitted to hospital will not appear in the STATS9 data, and vice versa. The chart shows different trends in STATS9 serious injuries and hospital admissions from the mid-99s, though since 25/6 both series have shown similar falls. Given the differences in definitions, data collection methodologies and use of the two datasets, interpretation of the trends shown is not straightforward, and there are a number of reasons why they may differ. As well as definitional differences, these could include: Changes in hospitals practices or how they record their data, particularly changes to the comparatively new HES system over time A change in the proportion of road casualties admitted to hospital A change in the number of less severe, non-hospitalised casualties which are still classed as serious in STATS9 many such cases will be handled in A&E only, and therefore not be recorded in the HES inpatient statistics Changes in the police recording of injury severity A change in the level of reporting of accidents by the public to the police. 4 See definitions section of this publication. 5 Finished in year emergency admission episodes, excluding non traffic accidents and in-hospital deaths were selected for this analysis. It should be noted that as HES contains one record for each period of care under a particular consultant, it is possible for a patient to be counted more than once (e.g. if they transfer to another consultant). 64

69 In our 26 report 6, we published an article comparing STATS9 and HES data which looked at the first point in the list above and concluded that: The overall increase in road traffic hospital admissions is accounted for by increases in admissions of less than 2 days, or unknown duration. This reflects changing hospital practices, for example the use of assessment or short-stay admission wards for monitoring. It is likely that the road casualty admissions data are being driven by changes affecting all admissions from A&E, for example the introduction of Payment by Results which has increased the importance of HES to the NHS. There have also been improvements to the coding of the HES data since 996. Thus, there is evidence that the increase in road traffic hospital admissions does not necessarily equate to an actual rise in the real number of road traffic casualties. However it is likely that the difference in the trends shown will be the result of a number of factors, and the extent to which each contributes is not fully understood. In order to investigate further, the Department has carried out matching of STATS9 and HES data at individual record level (summarised below). Table 5a summarises the number of emergency hospital admissions for 2 or more days (less likely to be affected by changes in admissions practices, though still influenced by other changes in hospital practices) and STATS9 seriously injured, broken down by age group and road user type. Note that the two data sources are not directly comparable. For example, the hospital data include admissions following falls from pedal cycles (often by children) which are not generally reported to the Police (see article 6 in Road Casualties Great Britain 26 for further details). However, they do show some similar trends over recent years. Both data sources show falls in the number of pedestrian, motorcyclist and car occupant casualties, over the last five years (between 23/4 and 27/8). Over the same period both sources also show a bigger reduction in child casualties, compared with adults. Overall the HES data shows a reduction in admissions (of 2 days or more) of 3 per cent between 23/4 and 27/8. This was marginally less than the overall fall in emergency admissions over this period, but a greater decrease than seen for overall injury admissions. It compares with a reduction of 6 per cent in seriously injured casualties reported to police over the period (which also includes some who are not admitted to hospital)

70 Table 5a: STATS9 and HES figures for England: 23/4 to 27/8 financial years Please note: figures are not directly comparable (see text) Numbers (thousands)/percentage 23/ 24/ 25/ 26/ 27/ Change /4 to 7/8 Hospital Episodes Statistics: emergency admissions for spells of 2 or more days () All admissions,584.9,543.3,469.7,394.2, All injury admissions (2) All Road traffic admissions (3) Pedestrians Pedal cyclists Motorcyclists Car occupants Age Age Age 65 and over STATS9 Seriously injured casualties All road users Pedestrians Pedal cyclists Motorcyclists Car occupants Age Age Age 65 and over () Finished inpatient admission episodes excluding in-hospital deaths (2) Episodes with an external cause of injury recorded (3) Episodes coded as a road traffic accident. Source: STATS9 and Hospital Episode Statistics (HES), The NHS Information Centre for Health and Social Care Matching HES inpatient data with STATS9 As noted above, there are many possible reasons why the STATS9 and HES datasets may show different trends, and comparisons of aggregate level figures do not provide a full insight into them. In order to gain a deeper understanding of the reasons for differences, the Department commissioned the Office for National Statistics to match the two datasets at individual record level. Although such matching does not enable an estimate of all road casualties to be made (as, for example, there are many that will not appear in either source), it provides some information on the number of hospital casualties not known to police, and vice versa. However, this is limited by the quality of the matching (which is highly dependent on the extent of postcode recording in the police data) 7. Table 5b shows the numbers of matched and unmatched records; last year s report contains further details of the matching methodology 8. We are currently reviewing this methodology following peer review and the figures are therefore subject to change. 7 It should be noted that as the matching process is not straightforward, there will be some records that are incorrectly matched, and some cases where matches have been missed. It is assumed that the effect of each is similar, but further work is required to establish whether this is the case. 8 See article 6 in Road Casualties Great Britain 27 for further details ( casualtiesgbar/) 66

71 Table 5b: Results of matching STATS9 and HES data for England: 999 to 24 Numbers (thousands)/percentage to 24 STATS9 serious Matched records Total records % matched STATS9 slight Matched records Total records % matched ,422 7 STATS9 all injuries Matched records Total records % matched ,69 Hospital road transport admissions () Matched records Total records % matched () The total number of records relates to files provided for matching by the Health and Social Care Information Centre, and include all road transport accidents, including those recorded as non-traffic accidents. Some cleaning of the data was carried out prior to matching and this means that totals will be different from figures published from HES data. Source: STATS9 and Hospital Episode Statistics, The NHS Information Centre for Health and Social Care The tentative conclusions from the work that we have done so far include: The proportion of road accident casualties admitted to a hospital that were known to police remained relatively constant between 999 and 24. There is no evidence that there was a systematic change in levels of reporting of serious accidents to police over this period, though these results are not conclusive either way, as the results could be affected by other factors (such as quality of data available for matching). There is, however, some evidence that the proportion of casualties admitted to hospital and known to police that were misclassified by the police as slightly injured increased marginally between 999 and This could be due to changes in police recording of severity, or changes in hospital practices, or a combination of both factors. Although the work done cannot conclusively answer the question of whether there has been a change in the level of reporting of accidents to police, the resulting matched dataset will be a useful source of information that can be used to compare the medical consequences of accidents (available in HES) with detailed accident circumstances recorded in STATS9. Article 6 presents an illustrative example of the sort of analyses that the matched data will facilitate. Acknowledgement We are grateful to the Health and Social Care Information Centre for allowing us to access the HES system. Copyright # 29, Re-used with the permission of The Health and Social Care Information Centre. All rights reserved. 67